Fourth

Edition

Simulation Modeling with SIMIO:

A Workbook

Jeffrey Allen Joines

Stephen Dean Roberts
North Carolina State University

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 SIMULATION MODELING WITH SIMIO:
A WORKBOOK V4

Jeffrey Allen Joines

Stephen Dean Roberts

North Carolina State University

Raleigh, North Carolina

October 2015

4th Edition

Copyright 2015 by Jeffrey A. Joines and Stephen D. Roberts All rights reserved.

Published by:
SIMIO LLC
504 Beaver St, Sewickley, PA 15143, USA
http://www.simio.com

SIMIO is a trademark of SIMIO LLC.

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Table of Contents
About the Authors ....................................................................................................................................... v
Preface to the Fourth Edition ....................................................................................................................vi
PART 1: ORGANIZATION OF THIS EDITION ............................................................................................... VI
PART 2: SPECIFIC CHANGES IN THE FOURTH EDITION ............................................................................ VII
PART 3: STYLES USED IN THIS BOOK ..................................................................................................... VIII
PART 4: ACKNOWLEDGEMENTS ............................................................................................................. VIII
Chapter 1 Introduction to Simulation: The Ice Cream Store ................................................................. 1
PART 1.1: WHAT IS SIMULATION? ............................................................................................................. 1
PART 1.2: SIMULATION FUNDAMENTALS: THE ICE CREAM STORE ........................................................... 3
PART 1.3: MANUAL SIMULATION ............................................................................................................ 10
PART 1.4: INPUT MODELING AND SIMULATION OUTPUT ANALYSIS ........................................................ 14
PART 1.5: ELEMENTS OF THE SIMULATION STUDY.................................................................................. 19
PART 1.6: COMMENTARY ......................................................................................................................... 20
Chapter 2 Introduction to SIMIO: The Ice Cream Store ...................................................................... 21
PART 2.1: GETTING STARTED .................................................................................................................. 21
PART 2.2: THE ICE CREAM STORE ........................................................................................................... 23
PART 2.3: ENHANCING THE ANIMATION .................................................................................................. 26
PART 2.4: LOOKING AT THE RESULTS ...................................................................................................... 29
PART 2.5: COMMENTARY ......................................................................................................................... 31
Chapter 3 Modeling Distance and Examining Inputs/Outputs ............................................................. 32
PART 3.1: BUILDING THE MODEL ............................................................................................................ 32
PART 3.2: USING THE 3D WAREHOUSE.................................................................................................... 33
PART 3.3: EXAMINING MODEL INPUT PARAMETERS ............................................................................... 34
PART 3.4: EXAMINING OUTPUT ............................................................................................................... 35
PART 3.5: USING EXPERIMENTS............................................................................................................... 37
PART 3.6: INPUT SENSITIVITY .................................................................................................................. 40
PART 3.7: COMMENTARY ......................................................................................................................... 41
Chapter 4 More Detailed Modeling: Airport Revisited ......................................................................... 42
PART 4.1: CHOICE OF PATHS .................................................................................................................... 42
PART 4.2: CHANGING ARRIVAL RATE ..................................................................................................... 44
PART 4.3: STATE VARIABLES, PROPERTIES, AND DATA TABLES ............................................................ 46
PART 4.4: MORE ON BRANCHING............................................................................................................. 50
PART 4.5: WORK SCHEDULES .................................................................................................................. 51
PART 4.6: COMMENTARY ......................................................................................................................... 53
Chapter 5 Data-Based Modeling: Manufacturing Cell .......................................................................... 55
PART 5.1: CONSTRUCTING THE MODEL ................................................................................................... 56
PART 5.2: SETTING CAPACITIES ............................................................................................................... 58
PART 5.3: INCORPORATING SEQUENCES .................................................................................................. 59
PART 5.4: EMBELLISHMENT: NEW ARRIVAL PATTERN AND PROCESSING TIMES ................................... 61
PART 5.5: USING RELATIONAL TABLES ................................................................................................... 64
PART 5.6: CREATING STATISTICS............................................................................................................. 68
PART 5.7: OBTAINING STATISTICS FOR ALL PART TYPES ....................................................................... 70
PART 5.8: AUTOMATING THE CREATION OF ELEMENTS FOR STATISTICS COLLECTION .......................... 72
PART 5.9: COMMENTARY ......................................................................................................................... 76
Chapter 6 Assembly and Packaging: Memory Chip Boards ................................................................. 78
PART 6.1: MEMORY BOARD ASSEMBLY AND PACKING .......................................................................... 78
PART 6.2: MAKING THE ANIMATION REVEAL MORE INFORMATION....................................................... 84
Table of Contents i

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 PART 6.3: EMBELLISHMENT: OTHER RESOURCE NEEDS ........................................................................ 86
PART 6.4: CHANGING PROCESSING TIME AS A FUNCTION OF THE SIZE OF THE QUEUE.......................... 88
PART 6.5: CREATING STATISTICS ............................................................................................................ 90
PART 6.6: COMMENTARY ........................................................................................................................ 92
Chapter 7 Using SIMIO Processes .......................................................................................................... 93
PART 7.1: THE ADD-ON PROCESS ........................................................................................................... 93
PART 7.2: THE ADD-ON PROCESS TRIGGERS: ILLUSTRATING “ASSIGN” STEP ....................................... 94
PART 7.3: CREATING AN INDEPENDENT “REUSABLE” PROCESS ............................................................. 96
PART 7.4: COLLECTING TALLY STATISTICS ............................................................................................ 97
PART 7.5: MANIPULATING RESOURCES................................................................................................. 100
PART 7.6: TOKENIZED PROCESSES ........................................................................................................ 102
PART 7.7: COMMENTARY ...................................................................................................................... 104
Chapter 8 Working with Flow and Capacity: The DMV ................................................................... 105
PART 8.1: THE DMV OFFICE ................................................................................................................. 106
PART 8.2: USING RESOURCES WITH SERVERS ....................................................................................... 110
PART 8.3: HANDLING FAILURES ............................................................................................................ 111
PART 8.4: SERVER CONFIGURATION ALTERNATIVES ............................................................................ 114
PART 8.5: RESTRICTING ENTITY TO FLOW TO PARALLEL SERVERS...................................................... 116
PART 8.6: THE WAITING ROOM SIZE..................................................................................................... 122
PART 8.7: USING APPOINTMENT SCHEDULES........................................................................................ 124
PART 8.8: CONTROLLING THE SIMULATION REPLICATION LENGTH ..................................................... 130
PART 8.9: COMMENTARY ...................................................................................................................... 134
Chapter 9 The Workstation Concept: A Kitting Process ................................................................... 135
PART 9.1: THE KITTING PROCESS .......................................................................................................... 135
PART 9.2: SEQUENCE-DEPENDENT SETUP TIMES .................................................................................. 138
PART 9.3: SEQUENCE-DEPENDENT SETUP TIMES THAT ARE RANDOM ................................................. 141
PART 9.4: USING MATERIALS IN THE KITTING OPERATION .................................................................. 143
PART 9.5: RAW MATERIAL ARRIVALS DURING THE SIMULATION ........................................................ 146
PART 9.6: IMPLEMENTING A JUST-IN-TIME APPROACH ........................................................................ 148
PART 9.7: COMMENTARY ...................................................................................................................... 151
Chapter 10 Inventories, Supply Chains, and Optimization ................................................................ 152
PART 10.1: BUILDING A SIMPLE SUPPLY CHAIN ................................................................................... 152
PART 10.2: PROCESSING ORDERS IN THE SUPPLY CHAIN SYSTEM ....................................................... 156
PART 10.3: CREATING THE REPLENISHMENT PART OF THE SUPPLY CHAIN SYSTEM ........................... 158
PART 10.4: USING AN EXPERIMENT TO DETERMINE THE BEST VALUES............................................... 160
PART 10.5: USING SMORE PLOTS TO DETERMINE THE BEST VALUES ................................................ 161
PART 10.6: USING RANKING AND SELECTION TO DETERMINE THE BEST SCENARIO ........................... 163
PART 10.7: USING OPTQUEST™ TO OPTIMIZE THE PARAMETERS ........................................................ 165
PART 10.8: MULTI-OBJECTIVE AND ADDITIONAL CONSTRAINTS USING OPTQUEST™ ........................ 167
PART 10.9: COMMENTARY .................................................................................................................... 170
Chapter 11 Simulation Output Analysis ............................................................................................... 171
PART 11.1: WHAT CAN GO WRONG? ...................................................................................................... 171
PART 11.2: TYPES OF SIMULATION ANALYSES ..................................................................................... 172
PART 11.3: OUTPUT ANALYSIS.............................................................................................................. 172
PART 11.4: AUTOMATIC BATCHING OF OUTPUT ................................................................................... 177
PART 11.5: ALGORITHMS USED IN SIMIO BATCH MEANS METHOD .................................................... 179
PART 11.6: INPUT ANALYSIS ................................................................................................................. 180
PART 11.7: COMMENTARY .................................................................................................................... 185
Chapter 12 Materials Handling ............................................................................................................. 186
PART 12.1: VEHICLES: CART TRANSFER IN MANUFACTURING CELL .................................................. 186
PART 12.2: CART TRANSFER AMONG STATIONS ................................................................................... 191

ii Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 PART 12.3: OTHER VEHICLE TRAVEL BEHAVIORS (FIXED ROUTE AND FREE SPACE TRAVEL) ............ 195
PART 12.4: CONVEYORS: A TRANSFER LINE ........................................................................................ 197
PART 12.5: MACHINE FAILURES IN THE CELL ....................................................................................... 201
PART 12.6: SORTING CONVEYORS ......................................................................................................... 202
PART 12.7: COMMENTARY ..................................................................................................................... 204
Chapter 13 Management of Resources: Veterinary Clinic .................................................................. 205
PART 13.1: UTILIZING THE FIXED RESOURCE OBJECT .......................................................................... 205
PART 13.2: DIFFERENT RESOURCE NEEDS BASED ON DIFFERENT PATIENT TYPES............................... 208
PART 13.3: RESOURCE DECISION MAKING ............................................................................................ 213
PART 13.4: ADDING AN ADDITIONAL PROCESS ..................................................................................... 215
PART 13.5: CHANGING PROCESSING BASED ON ANIMAL TYPE AND VET SERVICING .......................... 217
PART 13.6: CHANGING THE RESOURCE ALLOCATION SELECTION ........................................................ 223
PART 13.7: COMMENTARY ..................................................................................................................... 224
Chapter 14 A Mobile Resource: The Worker ....................................................................................... 226
PART 14.1: ROUTING PATIENTS ............................................................................................................. 226
PART 14.2: USING A WORKER AS A MOVEABLE RESOURCE .................................................................... 231
PART 14.3: RETURNING TO THE OFFICE BETWEEN PATIENTS................................................................ 235
PART 14.4: ZERO-TIME EVENTS ............................................................................................................ 238
PART 14.5: HANDLING MULTIPLE VETS ................................................................................................ 240
PART 14.6: COMMENTARY ..................................................................................................................... 242
Chapter 15 Adding Detail to Service: A Bank Example ...................................................................... 244
PART 15.1: USING A WORKER AS A RESOURCE AND A VEHICLE ........................................................... 244
PART 15.2: HAVING THE BANKER ESCORT THE CUSTOMER TO THE DEPOSIT BOX .............................. 248
PART 15.3: USING THE TRANSPORT FUNCTION OF THE WORKER ......................................................... 253
PART 15.4: RESOURCE RESERVATIONS.................................................................................................. 256
PART 15.5: ANIMATED ENTITIES ........................................................................................................... 257
PART 15.6: DETAILED SERVICE: TASKS AND TASK SEQUENCES .......................................................... 260
PART 15.7: USING TASK SEQUENCES .................................................................................................... 262
PART 15.8: SOME OBSERVATIONS CONCERNING TASKS ....................................................................... 264
PART 15.9: COMMENTARY ..................................................................................................................... 269
Chapter 16 Modeling of Call Centers .................................................................................................... 270
PART 16.1: BUILDING THE SIMPLE MODEL............................................................................................ 271
PART 16.2: BALKING .............................................................................................................................. 274
PART 16.3: MODELING RENEGING OF CUSTOMER CALLS ..................................................................... 276
PART 16.4: OPTIMIZING THE NUMBER OF FIRST LINE TECHNICIANS .................................................... 279
PART 16.5: USING THE FINANCIAL COSTS AS THE OPTIMIZING OBJECTIVE .......................................... 283
PART 16.6: COMMENTARY ..................................................................................................................... 287
Chapter 17 Sub-Modeling: Cellular Manufacturing ........................................................................... 289
PART 17.1: MODEL OF ONE WORK CELL ............................................................................................... 289
PART 17.2: CREATING THE SUB-MODEL................................................................................................ 291
PART 17.3: CREATING A MODEL USING THE WORKCELL SUB-MODEL ................................................. 293
PART 17.4: ADDING TO THE WORKCELL OBJECT .................................................................................. 294
PART 17.5: EXPOSING RESOURCE AND CAPACITY PROPERTIES ............................................................ 296
PART 17.6: PASSING INFORMATION BETWEEN THE MODEL AND ITS SUB-MODELS .............................. 298
PART 17.7: COMMENTARY ..................................................................................................................... 300
Chapter 18 The Anatomy of Objects: Server ....................................................................................... 301
PART 18.1: A SIMPLE RESOURCE MODEL: WAREHOUSE PICKUP ......................................................... 301
PART 18.2: TAKING AN OBJECT APART TO FIGURE OUT HOW IT WORKS.............................................. 303
PART 18.3: SIMIO OBJECTS AND CLASS HEIRARCHY ........................................................................... 308
Chapter 19 Building New Objects via Sub-Classing: A Delay Object................................................ 310
PART 19.1: SUB-CLASSING THE TRANSFERNODE TO CREATE A DELAYNODE .......................................... 310
Table of Contents iii

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 PART 19.2: MODIFYING PROCESSES AND ADDING A PROPERTIES FOR THE NEW NODE ....................... 312
PART 19.3: CREATING A MODEL TO TEST THE NEW DELAYTRANSFERNODE .......................................... 314
PART 19.4: COMMENTARY .................................................................................................................... 318
Chapter 20 Creating New Objects ......................................................................................................... 319
PART 20.1: CREATING A SIMPLE DELAY OBJECT .................................................................................. 319
PART 20.2: ADDING COLOR AND STATES .............................................................................................. 325
PART 20.3: ADDING DEFINED ADD-ON PROCESS TRIGGERS ................................................................. 327
PART 20.4: EMBELLISHING WITH STATE ASSIGNMENTS ....................................................................... 330
PART 20.5: CAPACITY AND ITS COMPLICATIONS .................................................................................. 334
PART 20.6: ADDING MULTIPLE RESOURCES ......................................................................................... 336
PART 20.7: USING STORAGES TO DISTINGUISH WAITING VERSUS DELAYING ..................................... 338
PART 20.8: SOME OBSERVATIONS ON THE DESIGN OF OBJECTS ........................................................... 341
PART 20.9: COMMENTARY .................................................................................................................... 341
Chapter 21 Continuous Variables, Reneging, Interrupt, Debugging: A Gas Station ...................... 343
PART 21.1: SIMPLE TANK MANUAL PROCESS ....................................................................................... 343
PART 21.2: SIMPLE TANK REVISITED USING THE FLOW LIBRARY ........................................................ 347
PART 21.3: THE GAS STATION............................................................................................................... 350
PART 21.4: RENEGING THE CARS WHEN THE PUMP GOES OFF............................................................. 356
PART 21.5: INTERRUPTING THE CARS WHEN PUMP GOES OFF ............................................................. 358
PART 21.6: USING ENTITIES THAT CARRY A TANK ............................................................................... 359
PART 21.7: COMMENTARY .................................................................................................................... 363
Chapter 22 More Subclassing: Advanced Modeling of Supply Chain Systems ............................... 364
PART 22.1: DEVELOPING A SPECIALIZED SUPPLY CHAIN WORKSTATION OBJECT............................... 364
PART 22.2: ADDING THE ORDERING STATION AND CHARACTERISTICS TO HANDLE ORDERS ............. 366
PART 22.3: ADDING THE BEHAVIOR LOGIC FOR THE ORDERING SYSTEM ............................................ 370
PART 22.4: ADDING THE BEHAVIOR LOGIC FOR THE INVENTORY REPLENISHMENT SYSTEM .............. 373
PART 22.5: USING THE NEW TO MODEL THE COMPLEX SUPPLY SYSTEM ............................................ 376
PART 22.6: ADDING A SECONDARY SUPPLIER FOR OVERFLOW ORDERS .............................................. 379
PART 22.7: COMMENTARY .................................................................................................................... 382
Chapter 23 More Subclassing: Process Planning/Project Management ........................................... 383
PART 23.1: PROCESS PLANNING ............................................................................................................ 383
PART 23.2: CREATING A SPECIALIZED TIMEPATH TO HANDLE ACTIVITIES .......................................... 384
PART 23.3: CREATING A JUNCTION OBJECT TO HANDLE PRECEDENT CONSTRAINTS .......................... 386
PART 23.4: CREATING SMALL NETWORK TO TEST THE NEW OBJECT .................................................. 392
PART 23.5: BUILDING THE EXAMPLE NETWORK ................................................................................... 394
PART 23.6: ADDING THE SLACK AND PERCENT OF TIME ON CRITICAL PATH CALCULATIONS ............. 395
PART 23.7: ADDING SLACK AND PERCENT OF TIME ON CP CALCULATIONS SECOND APPROACH ....... 398
PART 23.8: ERROR CHECKING TO MAKE SURE MODELER USES JUNCTION CORRECTLY ..................... 400
PART 23.9: COMMENTARY .................................................................................................................... 402
Appendix A Input Modeling .................................................................................................................. 403
PART A.1 RANDOM VARIABLES ............................................................................................................ 403
PART A.2 COLLECTING DATA ............................................................................................................... 403
PART A.3 INPUT MODELING: SELECTING A DISTRIBUTION TO FIT TO YOUR DATA ............................. 404
PART A.4 DISTRIBUTION SELECTION HIERARCHY ................................................................................ 409
PART A.5 EMPIRICAL DISTRIBUTIONS IN SIMIO .................................................................................. 409
PART A.6 SOFTWARE FOR INPUT MODELING ........................................................................................ 411
PART A.7 THE LOGNORMAL DISTRIBUTION.......................................................................................... 412
PART A.8 MODELING THE SUM OF N INDEPENDENT RANDOM VARIABLES.......................................... 413
Index ......................................................................................................................................................... 416

iv Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 About the Authors
JEFFREY A. JOINES is an Alumni Distinguished Undergradate Professor, an Associate
Professor and the Associate Department Head of Undergraduate Programs in the Department of
Textile Engineering, Chemistry, and Science at NC State University. He received a B.S. in
Electrical Engineering and B.S. in Industrial Engineering, a M.S in Industrial Engineering, and
Ph.D. in Industrial Engineering, all from NC State University. He received the 1997 Pritsker
Doctoral Dissertation Award from the Institute of Industrial Engineers. He is a member of IEEE,
IIE, ASEE, Tau Beta Pi, Etta Kappa Nu, Alpha Pi Mu and Phi Kappa Phi. His research interests
include supply chain optimization utilizing computer simulation and computational optimization
methods. Dr. Joines teaches graduate and undergraduate classes in computer information systems, computer based
modeling in Excel and VBA, Lean Six Sigma, and computer simulation modeling. Dr. Joines has also teaches
industry programs in the areas of Design for Six Sigma, Simulation and Six Sigma, and Data Management to Assist
in Six Sigma through the Textile Extension programs Six Sigma Black Belt and Master Black Belt program. Dr.
Joines served as the Program Chair for the 2005 Winter Simulation Conference (WSC) and the Proceedings Editor
for the 2000 WSC as well as developed and maintained the WSC paper management system from 2000-2009. He
currently serves on the WSC Board of Trustees representing the IEEE Systems, Man, and Cybernetics and was the
2015 Board Chair. He has also been an author and session chair for several Winter Simulation Conferences. He
received the 2014 Distinguished Service Award from INFORMS College on Simulation.
Dr. Joines is involved in utilizing technology in the classroom and how it impacts problem solving. He was awarded
the 2012 Alumni Distinguished Undegraduate Professor, the 2006 NC State University Outstanding Teaching
Award and College of Textiles Outstanding Teacher which allowed him to become a member of the Academy of
Outstanding Teachers. In 2009, Dr. Joines (along with Professor Roberts) was awarded the Gertrude Cox Award for
Innovative Excellence in Teaching and Learning with Technology for Transformative Large Scale Projects.

STEPHEN D. ROBERTS is the A. Doug Allison Distinguished Professor in the Edward P. Fitts
Department of Industrial and Systems Engineering at NC State University. Professor Roberts
received his: Ph.D., M.S.I.E., and B.S.I.E. (with Distinction) from the School of Industrial
Engineering at Purdue University. His primary teaching and research interests are in simulation
modeling and health systems engineering. He has been a faculty member at NC State University
since 1990, serving nine years as Department Head of the Department of Industrial Engineering
and three years as Interim Director of the Integrated Manufacturing Systems Engineering Institute.
Prior to serving at NC State, he was a faculty member in the Department of Internal Medicine at
the Indiana University School of Medicine and the School of Industrial Engineering at Purdue University as well as
the Director of the Health Systems Research Group at Regenstrief Institute for Health Care. Previously, he was a
faculty member in the Department of Industrial and Systems Engineering at the University of Florida and Director
of the Health Systems Research Division of the J. Hillis Miller Health Center, University of Florida. He has had
sabbaticals at Wolverine Software and the University of Central Florida/Institute for Simulation and Training.
Professor Roberts is a member of Alpha Pi Mu, Tau Beta Pi, Sigma Xi, Sigma Tau, and Omega Rho and a Fellow
of the Institute of Industrial Engineers. He has held Kaiser Aluminum Fellowship and a NDEA Title IV Fellowship.
He received the AIIE 1967 Graduate Research Award, Outstanding Paper Award at the 12th Annual Simulation
Symposium, the Pritsker and Associates Outstanding Undergraduate Teaching Award in the Purdue School of
Industrial Engineering, the CA Anderson Outstanding Teacher in the NCSU Department of Industrial and Systems
Engineering, the Outstanding Teacher from the NCSU College of Engineering, membership of the NCSU Academy
of Outstanding Teachers, the Gertrude M. Cox Award for Transformative projects from NC State University (with
Professor Joines), the Distinguished Service Award from INFORMS College on Simulation, and has served as
member, Vice- Chair, and Chair of the Winter Simulation Conference (WSC) Board of Directors representing
TIMS (now INFORMS) College on Simulation, and Secretary, Vice-President/Treasurer, President, and Past-
President of the WSC Foundation Board. He was the Proceeding Editor for the 1983 WSC, the Associate Program
Chair in 1985, and Program Chair in 1986. He has been a WSC presenter, session chair, track coordinator, and
keynote speaker.

Preface v

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Preface to the Fourth Edition
This edition of this workbook maintains the successful “participatory” style introduced in the first edition. You
don’t sit and read the book without a computer loaded with SIMIO (the book was created using version 7.124
of SIMIO™). We expect your active participation in using SIMIO as you turn the pages. We try to carry on a
conversation with you. Our belief is that simulation is not a spectator sport. You have to practice to gain skill
with it and you develop that skill through modeling practice. This book encourages you to practice and use
your skill, and feedback from earlier editions appear to validate the approach. This book retains its focus on
simulation modeling with SIMIO and most of the simulation statistical analysis and analytical issues are more
thoroughly covered in other books. We strongly suggest that if you are teaching/learning simulation that you
also have one of these non-language books available.1

We have deliberately tried to keep the price of the book low (i.e., the E-book or the paper copy). A relatively
new simulation language like SIMIO is constantly. In fact, the SIMIO developers have a history of new
releases (called “sprints”) about every three months. Any book that describes SIMIO will go out of date
quickly, so we have tried to track new features and update this book fairly often. If the book price is low,
maybe you will want to re-buy this book from time to time, so you have the latest information. Also, we feel
when learning and teaching, the paper copy allows the learner to write directly in the book.

Part 1: Organization of this Edition

This edition of the workbook has an evolved structure based on use and experience. More emphasis is placed
on “why” modeling choices are made, to supplement the “how” in using SIMIO in simulation. In Chapter 1,
we present fundamental simulation concepts, independent of SIMIO which can be skipped for those who
already understand these fundamentals. In Chapters 2 through 6, concentrates of the use of the Standard
Library Objects in SIMIO. You can do a lot of simulation modeling without resorting to more complex
concepts. A key part of those chapters is learning to identify/separate the data in a model from the model
structure. Chapter 7 introduces the fundamental topic of “processes,” which we frequently employ in the
following chapters. Chapters 8 and 9 concentrate on the important topics of flow and capacity. Chapter 10
introduces optimization in the context of supply chain modeling. Chapter 11 presents the influence of bias and
variability on terminating and steady-state simulation. Chapter 12 introduces SIMIO materials handling
features. Chapter 13 extends the use of resources while Chapters 14 and 15 describes the use of workers
including the detailed services provided by task sequences and their animation. Chapter 16 details the
simulation of call centers with reneging, balking, and cost optimization. Chapters 17 through 20 presents
object-oriented simulation capabilities in SIMIO. Chapter 17 builds a model out of an existing model (we call
it sub-modeling). Chapter 18 describes the anatomy of an existing SIMIO and in Chapter 19 we build a new
object by “sub-classing” an existing object. In Chapter 20 a new object is designed and built from a base
SIMIO object and its creation is contrasted with standard SIMIO object. Chapter 21 presents some of the
continuous modeling features in SIMIO. Chapters 22 and 23 demonstrates the power of object-oriented
simulation in the modeling supply chains and process planning respectively. We include an appendix on input
modeling, although SIMIO does not provide software.

The book is designed to be read from chapter to chapter, although it is possible to pick out certain concepts
and topics. Some redundancy is helpful in learning. By the time you have finished this book you should be
well-prepared to build models in SIMIO and to understand the virtues of different modeling approaches.

1
For example, Discrete-Event System Simulation (5th Edition), Jerry Banks, John Carson, Barry Nelson, David Nicol, Prentice-Hall,
2010 (622 pages)
vi Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Like SIMIO itself, this workbook has been designed for a variety of student, teacher, and practitioner
audiences. For example, if you are interested in manufacturing, you will want to be sure to study data-based
modeling in Chapter 5, assemply and packaging in Chapter 6, the workstation in Chapter 9, and material
handling in Chapter 12. If you are interested in logistics, don’t miss modeling of distances in Chapter 3, flow
and capacity in Chapter 8, inventories and supply chains in Chapter 10, and free space travel in Chapter12. If
you are interested in healthcare, be sure to review scheduled arrivals in Chapter 8, resource decision making in
Chapter 13, mobile workers in Chapter 14, and animated people and task sequences in Chapter 15. If object-
oreinted simulation is your interest, make sure to study Chapters 17 through 20, which describes how SIMIO
provides composition and inheritance to create objects. Manufacting examples and examples from the service
sector are used throughout. Also we pay some attention to input modeling (including input sensitivity) and
output analysis (including confidence intervals and optimization). This workbook provides comprehensive
and in-depth discussion of simulation modeling with SIMIO.

At the end of most chapters, we offer commentary on topics presented. We will emphasize the strengths and
weaknesses of the modeling approach and the language (we have no financial stake in SIMIO). To help
insure that everyone participates in this active learning process, we sprinkled questions throughout the
chapters. They have short answers and require the student pay some attention to what is going on. You can
use these in class. Accordingly, even though you don’t officially take attendance, you can give credit to
students who turn in their in-class assignment each day. These practices help develop a reputation as a class
you need to attend. We can provide you with answers to the questions, lecture notes, homework, and tests
through a shared Dropbox™ if you contact us.

Part 2: Specific Changes in the Fourth Edition

· In any type of software book, it is not until the book is actually used do mistakes crop up. This edition
corrects several mistakes and typos that occurred in the third edition.
· Since the third edition, SIMIO has gone through many changes and upgrades. The second edition and
to some extent the third edition just added the new features to existing chapters or added another
chapter later with the same case to highlight those new features (e.g., the original Chapter 7 which
introduced related tables based on the Chapter 4 case). In the new book we have combined those two
chapters which is now Chapter 5. We feel related data tables are an extremely important modeling
concept and needs to be introduced earlier in the development of the simulation model.
· Also, we moved all the material handling to a standalone chapter (i.e., Chapter 12). For those who
would like to have material handling sooner, Chapter 12 can be introduced any time after the Chapter
7 which introduces processes.
· As mentioned earlier, a new Chapter 1 which introduces simulation fundamentals which caused the
original Chapters 1 - 3 to shift to Chapters 2 – 4. Chapter 3 introduces the concept of Input Sensitivity
which is further explored in the new Chapter 11 devoted to Simulation Output Analysis. We do a more
thorough job of using confidence intervals to examine output.
· Chapter 6 is now the original Chapter 5 which introduced combiners for the memory chip problem.
However, the chapter is now followed directly by the processes chapter (i.e. Chapter 7) based on the
same memory chip problem rather than waiting several chapters as was done before.
· The new concept of Tokenized processes which allows one to create a generic process that can be
passed input parameters to specialize the process are an important concept. No longer do you have to
have almost identical processes. Tokenized processes are utilized in Chapters 7, 9, and 15.
· In the first three editions, the original Chapter 9 and 10 utilized a simple clinic to introduce concepts
of flow and capacity where chapter 9 was a little obsolete by the third edition. These two chapters
were combined into the new Chapter 8 improving the flow along with a different case (i.e., the DMV).
· Some chapters were just refreshed and updated to the new versions but have moved within the book.
For example, Chapter 9 and 10 are just refreshed versions of Chapter 8 and 11 from the third edition.
However, Chapter 9 does utilize Tokenized processes when producing the materials in the JIT section.
· Chapters 13 and 14 were combined into one chapter (Chapter 13) improving the flow. Another
example was added to the earlier editions when new features were added.
Preface vii

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 · Chapter 15 was updated to use the new Reserve option for acquiring resources. Also, SIMIO has
introduced more process capability beyond simple processing times by using the task sequences.
Chapter 15 explores the various features related to tasks utilizing the bank problem.
· Chapter 20 is now the new object chapter and was updated to include states and different animation
pictures along with capacity.
· Consistent with the third edition, the continuous chapter, Chapter 21 is located after the discussion of
objects. However, a new section was added that utilizes the new container entity, filler, and emptier
objects.
· The more advanced Chapters 22 (Shirt Folding Line), 25 (Lean Systems), and 26 (Multiple product)
were not included in this version for several reasons. Many of the new features of SIMIO have made
some of the tricks used since the first edition obsolete. Also we needed to limit the length of the book.
If anyone would like access to these chapters, we can provide you with the pdf and models of these
chapters not included.

Part 3: Styles Used in this Book

Certain styles have been used in this book to illustrate objects, names and parameters and to make it easier to
distinguish these types of parameters. Standard SIMIO objects will be set in small caps using a Courier New
font (e.g., SERVER) while objects that are created by the modeler will be also bolded (e.g., DELAYOBJECT).
Properties associated with these objects will be italicized (e.g., Processing Time). Process names will be
italicized and placed in quotes, as “OnEnteredProcessing”, while Add-on process triggers (e.g., Exited) will be
will be only italicized since they are properties. Process steps like Assign will be set in italicized in Courier
New font. SIMIO uses lots of expressions. These are set in Courier New font (e.g.,
SrvOffice.Contents >0). Names of all objects specified by the modeler will be bolded (e.g., Insert a
SERVER named SrvOffice). Values associated with properties will be set in a fashion similar to expressions or
in quotes for strings (e.g., “True”).

Part 4: Acknowledgements

We wish to thank our students, who have added much to our understanding and who let us often display our
ignorance. Also we thank the many colleagues and friends (new and old) who have read, commented as well
as used the book in their classes – hopefully they know who they are. In our writing this book, we appreciated
the response of SIMIO developers and SIMIO Support to our endless stream of questions and doubts,
especially Dave Sturrock and Dennis Pegden. Finally we thank our families for understanding and patience as
we often spend more time talking with each other than with them.

Please let us know how we can improve this workbook and how it can better meet your needs.

Jeff Joines ([email protected])

Steve Roberts ([email protected])
North Carolina State University

Raleigh, North Carolina

viii Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Preface ix

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Chapter 1
Introduction to Simulation: The Ice Cream Store
This chapter will give the novice in simulation an introduction to the terminology and mechanics of
simulation, as well as statistics used in performing a simulation model. If you are already familiar with these
basic simulation concepts, feel free to start with Chapter 2 which starts the simulation modeling in SIMIO.

Part 1.1: What is Simulation?

The word “simulation” has a variety of meanings and uses. Probably its most popular usage is in games. A
video game “simulates” a particular environment, maybe a formula one race, a battle, or a space encounter.
The game allows the user to experience something similar to what it is like to drive in a race, maybe with other
people participating. In the military, commanders create a battlefield simulation where soldiers act according
to their training perhaps defending a base or assaulting an enemy position. In aeronautical engineering, an
engineer may take a model airplane to a wind tunnel and test its aerodynamics. Most simulations have
elements of reality with the intention that the participant will learn something about the environment through
the simulation or perhaps only learning to play the game better.

We employ simulation to study and improve “systems” of people, equipment, materials, and procedures. We
use simulation to mimic or imitate the behavior of systems like factories, warehouses, hospitals, banks,
supermarkets, theme parks – just about anywhere a service is provided or an item is being produced. Our
simulations are different from gaming and training simulations in that we want the simulation to model the
real system, so we can investigate various changes before making recommendations. In that sense, simulation
is a performance improvement tool. The simulation acts as an experimental laboratory, except that our
laboratory is not physical but instead a computer model. We can then perform experiments on our computer
model.

Modeling
Many performance improvement tools (i.e., Lean, Six Sigma, etc.) rely on models. For example value stream
maps, spaghetti diagrams, process flow charts, waste walks, etc. are useful conceptual/descriptive models for
performance improvement based on direct observations of the system. These models provide a wide-range of
vehicles for describing and analyzing various systems. More formal models employ mathematical and
statistical methods. For example, linear regression is a popular modeling technique in statistics. Queuing
models offer a means for describing an important group of stable stochastic processes. Linear programming is
a formal optimization method of finding the values of variables that minimize or maximize a linear objective
function subject to linear constraints. Nevertheless, all these methods require a variety of assumptions about
the system being modeled. For example, the variables are constants (i.e., real or integer valued) and often
related linearly. If the variables have statistical variation, they are assumed to be normally or exponentially
(i.e., Markovian) distributed.

Simulation is a model-based improvement tool. However, few assumptions need to be made to build the
model. The model can be non-linear, described by arbitrary random variables, have a complex relationship,
and change with time (i.e., dynamic). In fact, the simulation model is limited only by your imagination and
the nature of the system being considered. You determine the nature of the model, based on what you think is
important about the system being studied.

Computer Simulation Modeling

Instead of a formal model, our simulations are computational. The model is essentially a logical description of
how the components of a system interact. This description is translated into a computational structure within
the computer using a simulation language. The computer simulation of the system is executed over and over
to generate statistical information about the system behavior. We use the statistics to describe the system
Introduction to Simulation: The Ice Cream Store 1

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 performance measures. Based on what we learn about the system, we modify the computer simulation models
to study alternative systems (i.e., experiment). By comparing these alternative systems statistically, we are
able to offer performance improvement recommendations.

Of course, we could just experiment directly with the system. If we thought an additional person on the
assembly line would improve its production, we could try that. If we thought a new configuration of the
hospital emergency room would provide more efficient care, we could create that configuration. If we thought
a new inventory policy would reduce inventory, we could implement it. But now the value of a computer
simulation model becomes apparent. A change in a computer simulation model is clearly cheaper and less
risky as compared to changing the real system. It is also faster to use a computer simulation to determine if
the changes are beneficial. Also, it might be safer to try a change in a simulation model than to try it in real
life. In general, it is much easier to try changes in a computer model than in a real operating system,
especially since changes disrupt people and facilities.

A computer simulation model allows us to develop confidence in making performance improvement

recommendations. We can try out a wide range of alternatives before disrupting an existing system. You can
now see why many companies and many managers require that a simulation study be done before making
substantial changes to any working system, especially when there are potential negative consequences of
performance changes as well as costs that don’t provide improvement.

Verification and Validation

Since simulation models often have serious consequence and are developed with a minimum set of
assumptions, the validity of a simulation model needs to be carefully considered before we believe in the
recommended benefits. In simulation, we often use the words “verification” and “validation” and we have
specific definitions for each as they relate to simulations.

The word “verification” refers to the model and its behavior. We most often develop our simulation models
with a simulation language. This language translates our modeling “intent” into a computation structure that
produces output statistics. Most simulation languages can be used to describe a complex operating system and
the language provides a framework of components for viewing that system. As our models become more
complex, we employ more complex simulation language constructs to model the behavior. This relationship is
critically dependent on our understanding of the simulation language intricacies. We may or may not fully
understand computational code and thus we confront the key question in verification: Does the simulation
model behave the way we expect? Is it possible that we may have made a mistake in employing the simulation
language or perhaps the simulation language creates an unexpected behavior in the computational code?
Therefore, even after we have created our simulation model, it needs to be tested to see if it behaves as
expected. If we make the processing time longer, does it result in longer times in the system? If we reduce the
arrival rate to a queue, does the waiting time decrease? The answers to any of these questions are specific to
the system and your model of it. But, above all else, we need to be sure that our model is behaving the way
we expect without error.

The word “validation” refers to the relationship between the model and the real system. Does the simulation
produce performance measures consistent with the real system). Now, we are assuming our model has been
verified but can we validly infer to the real system? This question strikes at the heart of our modeling effort
because without a valid model we cannot legitimately say much about performance improvement. Many
people new to simulation may want the model to be a substitute for the “real system” and that is a limitless
task. After all, the only “model” that is perfectly representative of the real system is the system itself!

We must always remember in simulation that our model is only an approximation of the real system. So the
most relevant way to validate our model is to concern ourselves with its approximation. How do we decide
what to approximate? It depends on our performance measures. If our performance measure is time in the
system, then we concern ourselves with those factors that impact time in system. If our performance measure
is production, then we focus on those factors that impact production. Usually we are interested in several
2 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 performance measures, but those measures will be the focus of our concern and will limit our modeling
activity. Otherwise without a clear set of performance objectives, we are left with a search for reality in our
model, which is a never-ending task.

Generally, people who have only a general familiarity with simulation, think simulation models can mimic
anything and often drive the development of a model that is needlessly complex. One of the difficult
responsibilities of anyone engaged in simulation modeling is the education of the stakeholders on the benefits
and limitations of a simulation model as well as the simulation modeling activity.

Computer Simulation Languages

The process of creating a computer simulation model varies from programming your own model in a
programming language such as C++, C#, or Java to using a spreadsheet like Excel. There are a variety of
simulation languages including Simula™, GPSS™, SIMSCRIPT™, Arena/SIMAN™, SLX™, ProModel™,
Flexsim™, AutoMod™, ExtendSim™, Witness™, AnyLogic, ™ among many others. Each of these
simulation languages differ in the way they require users to construct a simulation model. An important
distinction is the degree to which computer programming is required. Also many of the simulation languages
have evolved from a particular industry or set of applications and are especially useful in that context.

We have chosen to use the simulation language SIMIO which is a relatively new language, having been
developed over the past several years. The developers of the language had previously developed the Arena
simulation system. SIMIO benefits from more recent developments in object-oriented design and agent
modeling. SIMO is a “multi-modeling” language having agents as well as discrete event and continuous
language components. SIMIO was developed to provide visual appeal through its 3-D animation and
graphical representations. SIMIO provides a wide range of extensions, from direct modification of executing
processes to user developed objects. SIMIO also provides interoperation with various spreadsheets and
databases. Finally, SIMIO has become widely adopted in industry as well as academic institutions. Learning
SIMIO provides you with one broadly-based simulation modeling tool which can help you learn to use others
if the need arises.

Part 1.2: Simulation Fundamentals: The Ice Cream Store

A fundamental understanding of simulation will be beneficial throughout your study of simulation. It’s easy
to get caught up in the creation of a computer model using a simulation language and miss important basic
principles of simulation modeling. Too many people associate simulation with a simulation language and for
them, simulation is simply learning a simulation language. Learning a simulation language is necessary to
using simulation, but it does not substitute for understanding at a fundamental level. If you understand the
fundamentals of simulation, then you establish a basis for understanding any simulation language and any
simulation model. In fact, this understanding will be a key to learning almost everything else in this text.

The Ice Cream Store

It is helpful if the discussion of simulation is done in the context of a problem – albeit a simple problem.
Using this simple problem, we will be able to describe simulation elements of modeling, execution, and
analysis. Our problem is the common ice cream store since everyone loves ice cream and probably has visited
an ice cream store. As seen in Figure 1.1, customers arrive to the ice cream store where they obtain an ice
cream cone. It’s a simple store where there is only one attendant and people will wait in a single line to order
and receive their ice cream cone. The attendant waits on each customer, one at a time in the order they arrive.

Question 1: If you owned or managed the ice cream store, what might be your operational concerns?
________________________________________________________________________________

Likely one of your most prominent concerns would be this store’s operation and how you can improve it. For
example, could you buy a new cone making machine to make cones faster or should you hire someone to help
service customers? Should you resize the waiting space? These are performance improvement concerns.
Introduction to Simulation: The Ice Cream Store 3

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 1.1: The Ice Cream Store
Question 2: If you made one of these changes to the ice cream store, how would you decide if it improved
the store’s operation?
________________________________________________________________________________

In other words, what are the performance measures? Here are some possible measures: number of customers
served per day, time customers spent in the store, the waiting time of customers, the number of customer
waiting, and the utilization of servers. We will discover these are common performance measures and will
often be your first performance measurement choices.

Question 3: How can you expect to improve performance if you don’t know what is going on inside the
ice cream store?
________________________________________________________________________________

You need to employ all the descriptive tools you know to understand what is going on inside the system. At
this stage, one might create a value stream map, a flow chart, a relationship chart, a spaghetti diagram, etc.
These techniques will greatly improve your understanding of what happens to customers and the attendant
during the sale of an ice cream cone. Perhaps you develop the conceptual model (i.e., flow chart) presented in
Figure 1.2.

Customers Customer Served

Customer Exits
Arrive (capacity limited)
Possibly
Waiting

Figure 1.2: Flow Chart of the Ice Cream Store

The conceptual model shows how the customers are served. Notice, we have added the possibility of waiting
due to the fact that service is limited by the availability of the single attendant.

Gathering Data about the Ice Cream Store

Descriptive information helps us understand the service process in the ice cream store; however it doesn’t give
us the performance measures. For that we need to document what is happening (i.e., we need to do a “present
systems analysis”). Suppose we decide to do a “time study” of the store operations which is shown in Table
1.1.

4 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Table 1.1: Direct Observation Time Study (Event View) of Ice Cream Store
Time Customer Process
0 Store Opens – Server Idle
Arrives –
0 1
Start service on Customer 1 – Server Busy
8.36 1 Departs service – Server Idle
Arrives –
9.01 2
Start service on Customer 2 – Server Busy
9.98 3 Arrives – Customer 3 waits
13.5 4 Arrives – Customer 4 waits
2 Departs Service – Customer 2 leaves:
19.36
3 Start service on Customer 3 – Server Busy
23.07 5 Arrives – Customer 5 waits
3 Departs service – Customer 3 leaves:
27.22
4 Start service on Customer 4 – Server Busy
4 Departs service – Customer 4 leaves:
33.82
5 Start service on Customer 5 – Server Busy
38.18 6 Arrives – Customer 6 waits
40 End observations

In our time study, we simply record all the “events” and the event time that occurs in the ice cream store. An
event occurs when something operationally happens in the store, like an arrival of a customer or the departure
of a customer from service. Also we recorded when the attendant (i.e., server) becomes busy or idle. Here we
only observed the first 40 minutes of the store’s operation. We recognize this isn’t really enough time to gain
a full understanding, but we are not intending to solve a problem at this time – only to demonstrate a method.

The time study is sufficient for us to compute some performance measures, but first it will be helpful to re-
organize our time study data. We are not going to add (or subtract) any information. We are simply going to
re-organize our data from an event view to an entity (i.e., customer) view. The event view recorded events but
the entity view allows us to follow our customers. The re-organized data is shown in Table 1.2 which shows
when the customer arrives to the store, enters service, and leaves the store. Notice, customers five and six
service and exiting store happen after the 40 minutes.
Table 1.2: Re-organized Event Data in Customer View From
Customer Arrives to Store Enters Service Leaves Store
1 0.00 0.00 8.36
2 9.01 9.01 19.36
3 9.98 19.36 27.22
4 13.50 27.22 33.82
5 23.07 33.82 ??
6 38.18 ?? ??

Performance Measure Calculations

From the entity view of the time study data, we can easily compute a number of performance measures.

· Production: Number of people served

The number of people served is easily calculated by looking at the number of customers that have
exited the system in the simulation time. In our example, four people were served in 40 minutes or the
system had a production rate of six per hour (i.e., 4 customers/40 minutes * 60minutes/hour).
· Flowtime, Cycle Time: Time in System

Introduction to Simulation: The Ice Cream Store 5

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 The time in system is calculated by averaging the difference when the customer entered the system
from when they exited the system. Again, only four customers exited the system and contributed to the
average time in system in our simple ice cream system as seen in Table 1.3.
Table 1.3: Calculating Average Time in System for the Ice Cream Store
Customer 1 Customer 2 Customer 3 Customer 4 Average Time in System
8.36-0.0 19.36-9.01 27.22-9.98 33.82-13.5 56.27/4 = 14.07 minutes

· Non-Value Added Time: Waiting Time in Queue

The time in queue (i.e., the waiting or holding time) represents the time customers waited in line
before being seen by the ice cream attendant. It is calculated by averaging the difference when the
customer entered the system from when they entered service. During our observation window, five
customers entered service as seen in Table 1.4.
Table 1.4: Calculating Average Waiting Time in the Queue
Customer 1 Customer 2 Customer 3 Customer 4 Customer 5 Average Waiting Time
0.0-0.0 9.01-9.01 19.36-9.98 27.22-13.50 33.82-23.07 33.85/5 = 6.77 minutes

· Number Waiting on In Queue:

This measure determines the average number of customers one would expect to see waiting in line to
receive service. Unlike the previous metrics, number waiting in queue or number in the system
measures are time-persistent statistics or time-weighted statistics. You may have never computed a
time-persistent statistic which is a statistic for which we need to know the amount of time a value was
observed. We weight the value of the observation by the amount of time that value persists. A
graphical depiction of the number in the queue at the ice cream store is in Figure 1.3.

Figure 1.3: Graph of the Number in Queue in the Ice Cream Store
Consider, computing the average number in queue. Suppose we observed the number in queue to be
two one time and ten another time. Would you say that on average number we would expect to see six
in line? Of course “not”! You need to know how long the value of two was observed and how long
the value of ten was observed. Suppose we observe two in queue for ten minutes of the time and ten
in the queue for only one minute. So the queue was observed a total of 11 minutes. As a result, the
value of two was observed 10/11ths of the total time while ten was observed 1/11th of the total time.
Our average then is 2*(10/11) + 10 *(1/11) or = 2.73 customers. Or another way of computing is to
realize that the total waiting time2 observed was 20 + 10 over a total of 11 minutes which also yields
2.73 people. Now, looking at the data from the ice cream store, we need the percentage of time (40
minutes) that there was zero waiting, one waiting , two waiting , three and so forth3. In our case, a
maximum of two was observed and the time waiting would be computed as seen in Table 1.5.

2
Total waiting time is also the “area” under the curve in Figure 1.3.
3 We don’t include 0 since it makes no contribution to the total waiting time.
6 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Table 1.5: Calculating Average Number in Queue
Number In Queue Time Period Time Spent (min)
0 0.00 – 9.98 0*9.98 = 0.000
1 9.98 – 13.5 1*3.52 = 3.520
2 13.50 – 19.36 2*5.86 = 11.72
1 19.36 – 23.07 1*3.71 = 3.710
2 23.07 – 27.22 2*4.15 = 8.300
1 27.22 – 33.82 1*6.60 = 6.600
0 33.82 – 38.18 0*4.36 = 0.000
1 38.18 – 40.00 1*1.82 = 1.820
Total Waiting Time 35.67 minutes
Average Number in Queue 35.67/40 = 0.89 Customers

· Utilization:
This performance is the percentage time the server is busy servicing customers which is calculated by
dividing the time spent servicing customers divided by the time available. For our example, the
attendant is only idle during the time period between finish servicing customer one and the arrival of
customer two (i.e., 9.01 – 8.36 = .65 minutes). Therefore, the attendant’s utilization will be 39.35/40
or 98.4%:
· Other possible performance measures are maximum values, standard deviations, time between
departures, etc.
Table 1.6: Types of Performance Simulation Measures/Metrics
Type Description/Examples
Counts The number of parts that exited, entered, etc. (e.g., production).
These measures typically deal with time like waiting time, time in system
Observation-based Statistics
etc. SIMIO calls observation-based statistics “Tally” statistics.
Time-Persistence/Time- These measures deal with numbers in system, queue, etc. when the values
Average Statistics can be classified into different states. SIMIO calls these “State” statistics.

Question 4: What is another example of an observations-based performance measure?

________________________________________________________________________________

Question 5: What is another example of a time-persistence performance measure?

________________________________________________________________________________

Question 6: Is “inventory” a time-persistent or observation-based statistics?

________________________________________________________________________________

Question 7: Is the amount of time that a job is late a time-persistent or observation-based statistic?
________________________________________________________________________________

Terminology of a Queuing System

Figure 1.4 shows some elements of a “common” queueing system with the normal terminology. In general we
will refer to the arriving objects as “model entities” and the counter/attendant as a “server”. You can see the
members of the queue and the customer in service. The input processes for this system are the arrival process
and service processes. We will provide details on these inputs later.

Introduction to Simulation: The Ice Cream Store 7

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Ice Cream Store
4 Counter
Arriving 2 Departing
46 35
Customers Serviced Customers
(Model Entity)
Queue (FIFO) Customer in Service

Figure 1.4: A Breakdown of Terminology

Can we simulate it?

We want to be able to reproduce the time-study data collection exercise that we used for the ice cream store.
But we want to synthesize it numerically as opposed to observing it – in essence simulate it! If we can re-
create the time study data, we can compute the performance measures. Notice that the time study records are
centered on “events”. Recall these events are points in time when the system changes it state (i.e., status). A
quick review of that data reveals the following three events occur.

Table 1.7: Events of the Ice Cream Store

Number Type Event Description
1 Arrival Indicates an entity will arrive at this time.
2 End of Service An entity will be departing from service (i.e., service has finished)
3 End The simulation will terminate and end of all observations.
To facilitate our simulation we need a method to keep track of our events. Although there are other ways to
keep track of events, it is convenient to keep them in an “event calendar”. An event calendar contains the
records of future events (i.e., things we think are going to happen in the future), ordered by time (with the
earliest event first). Simulation can be executed by removing and inserting events into the event calendar.

Let’s add some specifics to our simulation problem (i.e., the ice cream store). We will use “minutes” as our
base measure of time. The input data that we currently have is shown in Table 1.8.

Table 1.8: Input Data

Customer Arrival Time Interarrival Time Processing Time
1 0.00 0.00 8.36
2 9.01 9.01 10.35
3 9.98 0.97 7.86
4 13.50 3.52 6.60
5 23.07 9.57 8.63
6 38.18 15.11 10.33
7 42.08 3.90 10.46
8 48.80 6.72 7.96
In the table, the arrival time has been re-stated as an “interarrival time” – namely time between arrivals.
Doing this re-statement does not change the arrival times but simply changes how they are presented. Such a
representation also requires the time of the first arrival from which the interarrival times are sequentially
computed. Table 1.8 provides data on only the first customers. We may need more data, but this is a later
discussion.
The Simulation Algorithm

To synthesize the system as presented in the time study, we need a systematic means of moving through time
by removing and inserting events on the event calendar. Consider the simulation algorithm shown in Figure
1.5.

8 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Algorithm Step 1: Setup the simulation (i.e., initialize the system)
Algorithm Step 2: Remove the next event from the event calendar
Algorithm Step 3: Update simulation time (i.e., TimeNow) to the time of that event
Algorithm Step 4: Execute the processes associated with the event (i.e., adding additional events
as needed and collecting statistics)
Arrival Event Process (see Figure 1.6)
Departure Event Process (see Figure 1.7)
End Event
Algorithm Step 5: Repeat Algorithm Steps 2-4 until complete
Figure 1.5: Simple Simulation Algorithm
Remember that the event calendar is ordered according to the next most recent event. So we can move
through time by removing the “next” event from the event calendar, updating time to the time of that event,
and executing whatever processes are associated with that event. This simple method is then repeated until we
reach some terminating time or condition.4 In our case, the “events” are the arrival of an entity, the service
(departure) of the entity, and the end of the simulation period. The arrival and departure of the entities are the
most important. Consider now how the arrival (see Figure 1.6) and the departure (see Figure 1.7) are
processed within the context of our simple single queue, single server system. The word “schedule” means to
insert this event into the event calendar.

Yes
Schedule Next Arrival Entity Goes Into
Arrival of Entity Is Server Busy Arrival Event Ends
(Create a New Event) Queue

No

Start Service
Resource
Schedule End of
Becomes Busy
Service Event

Figure 1.6: Event Process Associated with Arrival of Entity

No
Other Resource End of Service Event
End of Service Entity Exits System
Entities Waiting Becomes Idle Ends

Yes

First Entity in
Schedule End of
Queue Starts
Service Event
Service

Figure 1.7: Event Process Associated with End of Service

The arrival of an entity creates a new event because we know the next entity’s arrival time since we have the
interarrival times of entities. The newly arriving entity either must wait because the server (i.e., resource) is
busy or that entity can engage the server and start service. If the entity can start service, we now know a new
future event, namely the service departure because we know the processing time. So if the removed event is
an arrival, we insert a next arrival into the event calendar and may insert the service departure event in the
event calendar provided the entity can start service immediately. If the event removed from the event calendar
is a service departure, then the entity that has finished service will exit the system. If the waiting queue is
empty, then the resource (server) becomes idle. On the other hand, if there is at least one customer in queue,

4
The practice of removing the “next” event has caused some to refer to our simulation as a “next or discrete event simulation”.
Introduction to Simulation: The Ice Cream Store 9

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 then the first customer is brought into service and the resource remains busy and a new future event, namely
the service departure is inserted into the event calendar because we have the processing time. Note that a
service departure causes the entity to depart the system regardless of what happens to the server.

Question 8: What is the maximum number of new events that are added to the event calendar when an
entity arrival event occurs? What are they?
________________________________________________________________________________

Question 9: What is the minimum number of new events that are added to the event calendar when an
entity arrival event occurs?
________________________________________________________________________________

Question 10: What is the maximum number of new events that are added to the event calendar when an
entity departure event occurs? What are they?
________________________________________________________________________________

Question 11: What is the minimum number of new events that are added to the event calendar when an
entity departure event occurs?
________________________________________________________________________________

Finally, we note that the very first step of the simulation algorithm calls for the system to be “initialized.” In
other words, how will we start operation relative to the number of entities in line and the state of the server? It
will be convenient to use the “empty and idle” configuration. By “empty and idle” we mean that the server is
idle and the system is empty of all entities.

Part 1.3: Manual Simulation

To further understand how a discrete event simulation operates, a simple data structure will be employed to
execute a manual simulation as seen in Figure 1.8. Our manual simulation will consist of a “system animation”
graphical representation of what is going on in the system with the current customer being serviced shown
inside the square and other customers waiting outside the square. The current simulation time (i.e., called
“Time Now”) will be shown. The current event will be identified, followed by a description of the process.
Finally, the “Event Calendar” will be maintained, consisting of the event time and typ and the entity ID
number.

Figure 1.8: Manual Simulation Data Structure

Step 1: Algorithm Step 1 of the simulation as defined in Figure 1.5 is used to “initialize” the system as
seen in executing structure of Figure 1.9. Note, that we have inserted into the event calendar the time of
the first entity arrival and we have inserted the end event or the time to stop making observations.

10 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 1.9: System Initialization
Step 2: Executing the simulation algorithm, we remove the next event (Algorithm Step 2) from the event
calendar, update simulation time to the time of that event (Algorithm Step 3), and execute the appropriate
processes (Algorithm Step 4). The next event is the “Arrival of Entity #1” at time 0.0. The arrival of this
entity allows us to schedule into the Event Calendar the “Arrival of Entity #2” to occur since the
interarrival time between Entity #1 and Entity #2 is 9.01 minutes and thus the Arrival event for Entity #2
is at time 9.01 (i.e., 0.0 + 9.01). Because Entity #1 arrives when the server is idle, that entity enters
service and we can now schedule its Departure event because we know the processing time for Entity #1
as 8.36 minutes thus the event time is 8.36 (i.e., 0.0 + 8.36). The result is that our data structure now
appears as Figure 1.10.

Figure 1.10: Processing Time 0.0 Event

Question 12: We inserted the arrival of Entity #2 event before the departure of the Entity #1 event, so why
is the departure event ahead of the arrival event in the event calendar?
________________________________________________________________________________

Question 13: Why do we compute the new event times as 0.0 +? What does the 0.0 mean?
________________________________________________________________________________

Step 3: We are done with the processes associated with time 0.0. Now, the next event in the event calendar
is the “Departure of Entity #1” from the system at time 8.36. Since there are no entities in the queue, the
server is allowed to become idle and no new events are added to the event calendar as seen in Figure 1.11.

Figure 1.11: At Time 8.36

Step 4: The next event is the arrival of Entity #2 at time 9.01. So we remove it from the event calendar and
update the simulation time to 9.01. We can schedule the arrival of Entity #3 at the current time (9.01) plus
the interarrival time from Entity #2 to Entity #3 (0.97) which is event time 9.98. Next, Entity #2 can start
service immediately since the server is idle, so its departure event can be scheduled as the current time
(9.01) plus the processing time for Entity #2 of (10.35), which yields an event time of 19.36. The new
status of the simulation is shown in Figure 1.12. We also now have added animation of the status of the
entities and the server.

Introduction to Simulation: The Ice Cream Store 11

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 1.12: At Time 9.01
Step 5: The next event occurs at time 9.98 which is the arrival of Entity #3. Its arrival allows us to
schedule the next arrival (i.e., Entity #4) at time 13.50. But now the arriving entity must wait which is
indicated in the “System Animation” section of the data structure in Figure 1.13.

Figure 1.13: At Time 9.98

Question 14: How did we get 13.50 for the arrival of Entity #4?
________________________________________________________________________________

Step 6: The next event is the arrival of Entity #4, which will schedule the arrival of Entity #5, but will have
no other actions since the server remains busy. The new simulation time is 13.50 and the updated status is
shown in Figure 1.14

Figure 1.14: At time 13.50

Step 7: Finally, we see that Entity #2 finishes service at 19.36 and departs. Entity #3 can go into service
and we can schedule the service departure of Entity #3 at 27.22 as shown in Figure 1.15

Figure 1.15: At Time 19.36

Question 15: How did we compute the service departure for Entity #3 to be 27.22?
________________________________________________________________________________

Question 16: What is the next event?

________________________________________________________________________________

12 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 17: What new events are added as a result of this event? (Give the event time, the entity, and the
event type?
________________________________________________________________________________

Step 8: From Figure 1.15, we see the next event is the arrival of Enity #5 at time 23.07 which triggers the
addition of the next arrival (Entity #6).

Figure 1.16: At Time 23.07

Step 9: Entity #3 completes service at 27.22 which allows Entity #4 to enter serviced which schedules the
end of service event for Entity #4 at 33.82 as shown in Figure 1.17.

Figure 1.17: At Time 27.22

Question 18: What is the next event?
________________________________________________________________________________

Question 19: What new events are added as a result of this event? (Give the event time, the entity, and the
event type)?
________________________________________________________________________________

Step 10: The result of this next event is shown in Figure 1.18.

Figure 1.18: At Time 33.82

Question 20: What is the next event?
________________________________________________________________________________

Question 21: What new events are added as a result of this event? (Give the event time, the entity, and the
event type?
________________________________________________________________________________

Step 11: We are getting closer to the time to quit observing the system; however we have one more event
and one more time update. The result of this next event is shown in Figure 1.18.

Introduction to Simulation: The Ice Cream Store 13

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 1.19: At Time 38.18
Step 12: Finally, the next event calls for the “End” of the simulation at time 40.0. Only time is updated as
the event calendar is unchanged. The final state is given in Figure 1.20.

Figure 1.20: At Time 40.0

Our simulation approach is referred to as a “Discrete-Event Simulation” because the system only changes
states at defined event times and the state changes are discrete in that the number in queue changes discretely
as well as the number in system. If our model included variables that changed continuously with time, say like
the water in a tank, then we wouldn’t have a discrete-change system and we would have to consider all points
in time, not just those when the system changes state. Simulations that contain continuous variables are called
“Continuous Simulations.” And we can have combinations of the two types of simulations. In fact, SIMIO
can model both kinds of systems together – a multi-method simulation language.

Part 1.4: Input modeling and Simulation Output Analysis

Since we now have simulated the ice cream store time study, we can re-organize this information from an
entity viewpoint, and compute the same performance statistics that we computed previously!

Question 22: Do we have enough information to draw conclusions about the present system?
________________________________________________________________________________

Question 23: What can we do to extend the information available?

________________________________________________________________________________

If we had a total of 45 interarrival times and 43 processing times, we could simulate a total of 480 minutes of
time. For example, we may obtain the information given in Table 1.9. We have included the minimum,
maximum, and average values over the 480 minutes. Currently we simulated “only” one day of 480 minutes.

Question 24: Now do you have enough information for a “present systems analysis?
________________________________________________________________________________

Question 25: Is a one-day of simulation long enough?

________________________________________________________________________________

14 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Table 1.9: Final Performance Measures for 480 Minutes
Performance Measure Value
Total Production 43
Average waiting time in queue 9.59
Maximum waiting time in queue 35.65
Average total time in system 17.87
Maximum total time in system 42.65
Minimum total time in system 7.46
Time-average number of parts in queue 0.88
Maximum number of parts in queue 4
Ice Cream Attendant utilization 78%
One day doesn’t give us much information about the day so more “days” of information is needed. But that
means more interarrival times and more processing times. If we did a simulation of ten days, we would need
approximately 450 interarrival times and 430 processing times. If that information came from an electronic
record, then getting more data may not be a problem. 5 But if our only alternative is time studies, then we will
need to observe 10 days and that may be a greater intrusion into the actual operation than we can expect.

An alternative to more data collection is for us to find a “model” of this input data (i.e., interarrival and
processing times). Perhaps we can use a statistical representation since processing times and interarrival times
are most certainly random variables. We can often match the processing times and interarrival times to
standard statistical distributions. First we make a histogram of the data we observed. Second, we try to pick
out a statistical distribution, like a Gamma, Lognormal, Weibull, or Pert, to match the data. In fact, there is a
wide range of software that can help with this undertaking. An example of this activity is shown in Figure
1.21.

Figure 1.21: Matching Observed Processing Time to a Standard Statistical Distribution

As a result of this input modeling, we can characterize a random processing time or a random interarrival time
by a statistical model. For example, we might use a Pert distribution which has three parameters as: Pert (6, 8,
12) for a processing time or maybe a Gamma distribution which has two parameters as: 6.3 + Gamma(3.1,
0.7) where the 6.3 is the “location” of the distribution’s origin.

5
An implication of this need for more data is that performance improvement should be one of the bases on which an information
system is designed. The information should not be limited to accounting and reporting, but also performance improvement.
Introduction to Simulation: The Ice Cream Store 15

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Once we characterize an input with a statistical model, there is simulation technology6 that allows us to
“sample” from that statistical model. Repeated sampling will statistically reproduce the input model, but more
importantly give us an “infinite” supply of input data. With that extended data, we can simulate many days of
operation. Having many days of simulation provides us with the opportunity to gain an estimate of the
precision of the average performance measures we compute. For example suppose we simulate our system for
five days. Using simulation terminology, we performed five “replications” or “runs” of our simulation model.
Each replication obtains new “samples” for its interarrival times and processing times. To compute the
average and the standard deviation, we use the averages from each day (one day yields one average). The
results are shown in Table 1.10.

Table 1.10: Results from Five Replication of 480 Minutes Each

Replication 1 2 3 4 5 Avg. Std.Dev.
Avg. time in queue 9.59 15.26 12.98 8.08 21.42 13.47 5.26
Avg. no. in queue 0.88 1.62 1.17 0.76 2.31 1.35 0.63
Machine Utilization 0.78 0.86 0.75 0.78 0.81 0.80 0.04
Question 26: How many observations are used to compute the average and standard deviation in Table
1.10?
________________________________________________________________________________

Question 27: Why only the averages instead of all the queue waiting times within a given replication?
________________________________________________________________________________

We know that these statistics, like the average and the standard deviation, are themselves random variables. If
we looked at another five days (either simulated or real) we wouldn’t get the exact same results because of the
underlying variability in the model. We need to have some idea about how precise these summary statistics
are. In order to judge how precise a given statistic is, we often use a confidence interval. For example, we
computed the average waiting time as 13.47, but does this estimate have a lot of variability associated with it
or are we pretty confident about that value?

Question 28: Confidence intervals in statistics are based on what famous distribution as well as famous
theorem?
________________________________________________________________________________

Any observations for which we want to create a confidence interval must satisfy the assumptions that the
observations are “independent and identically” distributed. As a consequence the Central Limit Theorem
(CLT) can be employed to compute a confidence interval.

Question 29: Are the observations, say of the waiting times, during a simulated day independent and
identically distributed?
________________________________________________________________________________

Question 30: For example, would the waiting time for entity #1, entity #2, and entity #3 be independent of
each other? Would you expect the distributions of these waiting times to be identical?
________________________________________________________________________________

Question 31: Would the number waiting in the queue at 9am be independent of the number waiting at
8:45am? Would the distribution of the number waiting be identical?
________________________________________________________________________________

6
Random number and random variate generation methods.
16 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Almost no statistic (performance measure)7 computed during a single simulation replication will be
independent and identically distributed (e.g., the waiting time of entity #4 maybe dependent on entity #3
waiting and processing times). So instead, we use, for example, the average values computed over the day as
an observation (i.e., we will determine the average of the averages).

Question 32: Are daily averages independent and identically distributed? Why?
________________________________________________________________________________

Since these daily averages (or maximums or others) are independent and identically distributed, we can
compute confidence intervals for our output statistics (based on the Central Limit Theorem).

Confidence Intervals on Expectations

A confidence interval for an expectation is computed using the following formula:

where is the sample mean, is the sample standard deviation of the data, is the upper critical

point from the Student’s t distribution with degrees of freedom, and is the number of observations

(i.e., replications). Note, is often referred as the standard error or the standard deviation of the mean. As the

number of replications is increased, the standard error decreases. So a 95% confidence interval of the expected
time in the queue using the data from Table 1.10 would be computed as the following calculation.

Question 33: Are you confident in the 13.47 average?

________________________________________________________________________________

Question 34: Would you bet your job that the “true” average waiting time is 13.47 minutes?
________________________________________________________________________________

So another way to express the confidence interval is given by the Mean ± the Half-width, as in

which SIMIO will use. The Central Limit Theorem allows the calculation of the confidence interval by
asserting that the standard error (or the Standard Deviation of the Mean) can be computed by dividing the
standard deviation of observations by the square root of the number of observations. If the observations are
not independent and identically distributed then that relationship doesn’t hold. As n increases sufficiently large
the CLT and the inferential statistics on the mean of the population become valid.

Question 35: Does the confidence interval we computed earlier (i.e., ) mean that 95% of the
waiting times fall in this interval?8
________________________________________________________________________________

7 Of course there are exceptional cases.

8
Prediction intervals are used for observations.
Introduction to Simulation: The Ice Cream Store 17

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 36: Does it mean that if we simulated 100 days that 95% average daily waiting times would fall in
this interval?
________________________________________________________________________________

Question 37: Or looked at another way. Can we say that there is a 95% chance that the “true, unknown”
overall mean daily waiting time would fall in this interval?
________________________________________________________________________________

So the confidence interval measures our “confidence” about the computed performance measures. A
confidence interval is a statement about the mean (i.e., the mean waiting time), not about observations (i.e.,
individual waiting times). Confidence intervals can be “wider” than one would like. However, in simulation,
we control how many days (i.e., replications) we perform in our analysis thus affecting the confidence on the
performance measures.

Question 38: Using simulation, how can we improve the precision of our estimates (reduce the confidence
interval width?
________________________________________________________________________________

So we can run more replications in our simulation if we want “tighter” confidence intervals.9

Comparing Alternative Scenarios

In simulation, we usually refer to a single model as a simulation “scenario” or a simulation “experiment”. Our
model of the present ice cream store is a single scenario. However using simulation as a performance
improvement tool, we are interested in simulating alternative models of, for example, our ice cream store. We
would refer to each model as a scenario. So, for example, if we added a new ice cream making machine, this
change would constitute a different simulation scenario. If we started an ad campaign and expected an
increase in business in the ice cream store, we would have yet another scenario. In general, we would
typically explore a whole bunch of scenarios, expecting to find improvements in the operations of the ice
cream store. The simulation is our experimental lab.

So let’s reconsider a different scenario for our ice cream store. What would happen if the arrival rate was
increased by 10% (i.e., more customers arrive per hour owing to an ad campaign)? We could reduce the
interarrival time by 10% and make five additional replications of the 480 hour day. The results of the original
and this alternative scenario are shown in Table 1.11 along with the confidence intervals.

Table 1.11: Results of Original and Added Time Scenarios

Replication 1 2 3 4 5 Avg. Std.Dev. LCL UCL
Avg. time in queue 9.59 15.26 12.98 8.08 21.42 13.47 5.26 6.93 20.00
Avg. no. in queue 0.88 1.62 1.17 0.76 2.31 1.35 0.63 0.56 2.13
Machine Utilization 0.78 0.86 0.75 0.78 0.81 0.80 0.04 0.74 0.85
Increased Customer Arrival
Replication 1 2 3 4 5 Avg. Std.Dev. LCL UCL
Avg. time in queue 14.77 24.42 18.39 12.10 31.18 20.17 7.69 10.62 29.72
Avg. no. in queue 1.50 2.77 1.80 1.18 3.74 2.20 1.05 0.90 3.50
Machine Utilization 0.85 0.92 0.82 0.82 0.85 0.85 0.04 0.80 0.90
Variability in the outcome creates problems for us. We recognize that we simply cannot compare scenarios of
only one replication, but even with five, it’s hard to know if there are any real differences (although it appears
so). Once again we need to rely on our statistical analysis to be sure we are drawing appropriate conclusions.

9
Note that we can also create a smaller confidence level by using is a larger significance level or α value)
18 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 When comparing different sets of statistics like this, we would resort to the Student’s t-test. A way to conduct
the t-test is to compare confidence intervals for each of the two scenarios. If the confidence intervals overlap
then we will fail to reject the null hypothesis that the mean time in system for the original system equals the
new system.

The 95% confidence interval for the original five days was [6.94, 20.00] while for the increased arrival rate,
the confidence interval is [12.48, 29.76]. Now comparing the confidence intervals, we see that they “overlap”,
meaning we cannot say there is a statistical difference in the average waiting time. Without a statistical
difference, any statement about the practical difference10 is without statistical foundation.

Question 39: What can we do to increase our chance of obtain a statistical difference?
________________________________________________________________________________

Question 40: It is necessary to have a statistical foundation for our recommendations?

________________________________________________________________________________

Although it is probably unnecessary to have a statistical foundation for every recommendation, we strive to do
so to avoid the embarrassing situation of making a claim that is later shown to be erroneous – especially since
our job may be at risk. By striving to have a statistical foundation for our recommendations, we take
advantage of our entire toolbox in decision-making and promote our professionalism.

Part 1.5: Elements of the Simulation Study

The entire simulation study is composed of a number of elements, which we present here. Although these are
presented in a sequential manner, rarely is a simulation study done without stopping to return to an earlier
issue whose understanding has been enhanced. In many instances we work on several of the elements at the
same time. But regardless of the order, we usually try to complete all the elements.

· Understand the system: Getting to know and understand the system is perhaps the most intense step and
one that you will return to often as you develop a better understanding of what needs to be done.
· Be clear about the goals: Try to avoid “feature or scope creep”. There is a tendency to continue to
expand the goals well beyond what is reasonable. Without clear goals, the simulation effort wanders.
During this phase identify the performance measures of interest that will be used to evaluate the
simulation.
· Formulate the model representation: Here you are clearly formulating the structure and input for your
model. Don’t spend a lot of time doing data collection at this point because, as you develop the model, its
data needs will become clearer. Also be sure to involve the stakeholders in your formulation, to avoid
missing important concerns.
· Translate your conceptual representation into modeling software, which in our case is SIMIO. A lot of
time is spent learning SIMIO so you have a wide range of simulation modeling tools with which to build
this and other models.
· Determine the necessary input modeling: At some point data will need to be collected on the inputs
identified during the formulation and translation of the system into a computer model. The initial
simulation model can be used to determine which inputs are the most sensitive with regard to the output
which need to be collected. Fitting distributions is generally better but expert opinion can be used to get
the model up and running.
· Verify the simulation: Be sure the simulation is working as it is expected without errors. Do some “stress
tests” to see if it behaves properly when resources are removed or when demand is increased. Explain any

10
A practical difference means that the difference is important within the context of the problem. When we are concerned with a
practical difference of unimportant or cheap items, then a statistical difference is not necessary.
Introduction to Simulation: The Ice Cream Store 19

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 “zeros” that appear in the output. Don’t assume you are getting counter-intuitive results when they may
just be wrong.
· Validate the model: How does the model fit the real world? Is the simulation giving sufficient behavior
that you have confidence in its output. Can you validly use the model for performance improvement?
· Design scenarios: Determine which alternatives you think will improve the performance of the present
system and create the alternative simulation models and associate the models with scenarios.
· Make runs: do the simulation experiments for the scenarios. Be sure your simulation output generates the
appropriate performance measures. Make multiple runs for each scenario.
· Analyze results and get insight: Examine carefully the output from the scenarios and begin to develop
judgments about how to increase the performance of the system. Be sure the statistical analysis supports
your conclusions.
· Make Recommendation and Document: Be sure to discuss the results with all the stakeholders and
decision-makers. Make sure you have addressed the important problems and developed feasible
recommendations. Document your work so you or someone else can return one year later and understand
what you have done.

Part 1.6: Commentary

If you have worked carefully through this chapter, you will have a fundamental understanding of simulation
that is completely independent of the simulation software or any particular application. Some of the key
points have been.

· A simulation model consists of a system structure and input.

· The insertion and removal of events drive a simulation.
· Random variables are used to represent input.
· Simulation statistics include observations and time-persistent values.
· A simulation may consist of many performance measures.
· Verification and validation are important concerns in any simulation.
· By using a confidence interval, we have some measure of variability as well as central tendency of the
simulation output.
· Computerization of simulation greatly facilitates its value in the present and in the future.

20 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Chapter 2
Introduction to SIMIO: The Ice Cream Store
Simulation is a very useful tool. Just about everyone who learns about simulation gets excited about its
widespread applicability. Practically any manufacturing, production, or service system can benefit from a
simulation-based investigation. SIMIO can make the application of simulation easier to do and at the same
time provide a powerful approach for addressing complex problems for designing and improving these types
of systems.

Also, simulation is fun when you build models that are visually appealing and SIMIO kicks the fun up a notch
by offering 3D (i.e., three dimensional) visualization. Not only does the animation allow you can see how the
model is behaving, but people you work with can as well which will greatly increase your credibility with
them.

Part 2.1: Getting Started

We are assuming you have installed SIMIO on your computer and are now ready to get going.

Step 1: Invoke SIMIO by either opening the Start menu in Windows or by clicking on a SIMIO icon
that perhaps you have located on the Windows taskbar or on the Desktop. The opening SIMIO window as
seen in Figure 2.1 has the standard Microsoft Office “look and feel”.

Step 2: When you open SIMIO for the first time, you will be placed into a modeling environment. The
“Run Tab” of the “SIMIO ribbon” has been selected and contains the Run, Run Setup, Animation Speed,
and Display sections.

Step 3: Below the SIMIO ribbon are the SIMIO window tabs. The “Facility” window has been selected
and is the canvas that will be used to build your simulation models. The SIMIO window consists of the
display of “Libraries” area on the left, a “Browse” section on the right, and the middle section is the
modeling canvas.

Step 4: The Libraries section displays the [Standard Library] which contains the modeling object
definition icons. Those definitions are used to create objects by clicking on the particular definition icon,
dragging the icon to the modeling canvas to position it, and then clicking to complete the object definition.
Most of the icons have names that connote their meaning. Also under the Libraries section is the [Flow
Library] which will show the object definition icons for modeling various “flow” characteristics. Finally,
the last portion of the Libraries section is the “Project Library” which will contain the objects that become
a part of your SIMIO simulation project including the MODELENTITY object.

Step 5: In the Browse section, the top section is the “Navigation” panel, which identifies the components of
your simulation model. By default, a new project will be named “MySimioProject” and will consist of
two defined objects, a “MODELENTITY” and a “MODEL”. The lower section of the navigation panel displays
the “Property Inspector”. This panel will display the properties of the object selected in the navigation
panel. A specification in the Property Inspector is to “Show Commonly Used Properties Only” which will
filter out certain properties. Since we want to learn all of SIMIO, we will not use this screening and
should uncheck the option.

Step 6: Finally, notice the “Help” options in the upper right-hand corner which can also be invoked
through the F1 key. There is lots of information to be found in the help documents, especially after you
have some experience with SIMIO concepts and features.

Introduction to SIMIO: The Ice Cream Store 21

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Ribbon Tabs Units Help

SIMIO
Ribbon

File
SIMIO
Tab MySimioProject
Run Tab Windows

Navigation
Panel Objects in
MySimioProject

Property
[Standard Inspector
Library]
Modeling
Canvas

[Project
Library]

Figure 2.1: Opening SIMIO Window

Other SIMIO resources available to you are found under the SIMO ribbon tab “Support”. In the “Learning
Simo” section, the “SimBits” will refer you to a library of elemental models (each illustrating a particular
modeling structure inside of SIMIO) and the “Examples” section will refer you to a library of complete models
that provide interesting SIMO models. And there are a variety of Books, Videos, Training, and guides. Also
note the SIMIO “version”11 in the “My Software” section. You will want to keep your version up-to-date.

Step 7: When a new model is created (see Figure 2.1), two objects are automatically defined. The
MODELENTITY will create entities that move through our model while the MODEL will contain positioned
objects and the flow paths of entities. In a sense, the entities will roam around the positioned or fixed
objects.

Step 8: Right-click the mouse on the MODELENTITY and then on the MODEL objects in the [Navigation]
Panel and select the “Properties”.

Question 1: What is the “default” Model Name of the MODELENTITY?

________________________________________________________________________________

11
SIMIO “Sprint” releases come out approximately monthly, so the software will be changing rather quickly. Be sure you obtain
updates with new features and possibly bug fixes.
22 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 2: What is the Object Type of the MODEL (look under the “Advanced Options”)?
________________________________________________________________________________

Step 9: Note the Undo/Redo buttons. These can be used to correct mistakes and recover previous modeling
components. When an action cannot be undone, SIMIO will issue a warning.

Step 10: The “Units” are a convenient way to set the default specification units for time, length, rate, area,
volume and mass.12

Part 2.2: The Ice Cream Store

A small ice cream store sells ice cream cones. Customers arrive and wait in line to be served by one of two
attendants. These attendants take the ice cream order and give the cone(s) back to the customer who then
moves to the separate cashier to pay. After paying, the customers will leave the store. In building a
simulation model, it is often important to flowchart the processes. Figure 2.2 shows the four processes of the
ice cream store.

Customers 15 s Order Ice 15 s Pay for Ice 15 s Customer

Arrive Cream Cone Cream Leaves
Two Attendants One Cashier
Exponential
Triangular (0.4, 0.9, 1.5) Triangular (0.3, 0.4, 0.6)
(0.5) Minute
Minutes Minutes

Figure 2.2: Flowchart of Ice Cream Store

In this problem we will assume we know that:
· Customers arrive Exponentially with a mean interarrival time of 0.5 minutes,
· The time an attendant takes to interact with the customer and give them their ice cream cone(s) is
modeled with a Triangular distribution with a minimum of 0.4 minutes, a most likely time of 0.9
minutes, and a maximum of 1.5 minutes,
· The time the cashier takes to accept payment for the ice cream is also Triangular with a minimum of
0.3 minutes, a most likely time of 0.4 minutes, and a maximum of 0.6 minutes, and
· The travel time between each process is 15 seconds.

Customers will wait in a single waiting line on a first-come, first-served basis if both attendants are busy.
Likewise there is a single waiting line for the cashier. We will also assume there is no limit to the length of
the waiting lines.

Step 1: It should be fairly clear that the customers should be modeled by a MODELENTITY while the MODEL
will consist of the flowchart of the Ice Cream Store as seen in Figure 2.2. Entities are the objects that will
traverse through the network of fixed objects. Add an entity to your model by clicking on the
MODELENTITY object in the [Project Library] panel and dragging it onto the canvas.

Step 2: Add objects to your model by clicking on the object type in the [Standard Library] panel. Drag the
object around the modeling canvas and click to drop it into position (you can left-click and drag the
objects around to relocate or delete them from the model). When you click on an object, its properties
appear in the property inspector (typically on the bottom right side panel.

· Add a SOURCE object, two SERVER objects, and one SINK object – see Figure 2.3.

12
SIMIO internally keeps all times in hours, distances in meters, rates in per hours, weights in kilograms, areas in square meters and
volumes in cubic meters.
Introduction to SIMIO: The Ice Cream Store 23

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 · Connect the objects with TIMEPATH objects by clicking on the “output” node (blue diamond) of an
object and connecting it to the “input” node (grey diamond) of an object. You can click between the
nodes to produce a multi-segment path which allows the path to be more flexible.

Note the change in the ribbon

Model Entity

Source Object
Selected
Properties
Library
of Source

Server

Name of TransferNode
Object Output from
Objects
TimePath

BasicNode
Input to
Objects

Figure 2.3: Beginning to Model

Step 3: Change the defalut “names” for each object. You can change the name on the object directly13 or
in the “General” property section of for each object. Use the names as specified in Figure 2.3.
Question 3: A SERVER object has three “lines” that surround it. What are they called? (note that a “.” adds
further specification)
________________________________________________________________________________
Question 4: A SERVER has an “input” and an “output” node. These nodes are also [Standard Library]
objects. What is the standard library name of the input node?
________________________________________________________________________________

Question 5: What is the standard library name for the output node?
________________________________________________________________________________

Step 4: Next, click on each object and fill out its properties according to our assumptions. Figure 2.4 shows
the properties associated with the SOURCE object.14

13
Double click or right-click the object to change the name the directly. Do not forget to click enter to save the name change.
24 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Source Object Name of Entity
Properties

Arrivals based on
Interarrival Time

Time units
Expression Editor

Reveal Property
Details Description

Figure 2.4: Properties of Source

It is important to be sure the time units are correct for expressions involving time and distance. You may need
to reveal the details in a dialog by clicking the icon to show the additional information. The Expression
Editor allows you to write expressions which have the following form.

Step 5: Object.SubObject.SubObject(parameters)

To begin the expression editor start by specifying a character. Select the object, its subobject(s), and its
properties. The expression editor has “tab-completion” which means the tab key will complete the name from
its first few characters. It also indicates if subobjects are available for this object.

Question 6: What units of time are available for the interarrival time?
________________________________________________________________________________

Step 6: Complete your modeling by adding the information for the two SERVERS and TIMEPATHS:

· For the SrvGetIceCream server object:

o Initial Capacity: 2
o Processing time: Random.Triangular(0.4, 0.9, 1.5)
o Units: Minutes

· For the SrvPayCashier server object:

o Processing time: Random.Triangular(0.3, 0.4, 0.6)
o Units: Minutes
· For the TIMEPATHS objects:
o TravelTime: 15
o Units: Seconds

14
Note the default properties that are changed will be bolded to indicate a changed has occurred.
Introduction to SIMIO: The Ice Cream Store 25

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Note that the object names are listed under the “General” category in the property inspector.

Step 7: Before we run the model, let’s change the ending time in the Run Setup section of the SIMIO
“Run” tab to “Unspecified (Infinite)” as seen in Figure 2.5.

Set Ending
time

Figure 2.5: Setting the Length of the Run

Step 8: Click the run button to start the simulation and let the simulation “run” for a while.

Step 9: Notice the MODELENTITIES ( ) as they move through the model across the TIMEPATHS. The
entities queue in the InputBuffer.Contents and they are in service in the
Processing.Contents, both shown as “green lines” that surround the SERVER objects. You can
relocate the objects while the model is running.

Step 10: Click the “View” tab and select “3-D”.

Question 7: Hold down the left mouse button key and move the mouse left and right and up and down.
What happens?
________________________________________________________________________________

Question 8: Hold down the right mouse button key and move the mouse left and right and up and down.
What happens?
________________________________________________________________________________

Step 11: Switch between 2-D and 3-D using the “2” and “3” as “hotkeys”

Question 9: What happens in 2-D when you hold down the left mouse key and move the mouse left and
right and up and down?
________________________________________________________________________________

Question 10: What happens in 2-D when you hold down the right mouse button key and move the mouse
left and right and up and down?
________________________________________________________________________________

Step 12: Experiment in 2-D and 3-D by moving stations and changing the “layout” of the model. Change
the travel time to the SrvGetIceCream station to 50 seconds to see another kind of change.

Part 2.3: Enhancing the Animation

Animation is often what gets people interested in a model. A SIMIO 3-D model, although it might take some
work, can bring a lot of attention. In this section, we will give you some instructions on how to create an
animation that appears as in Figure 2.6.

26 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 2.6: The Animation
Step 1: Make the ice cream store customers look like people, rather than triangles. Select the
EntCustomer entity (i.e., the green triangle) and under the Symbol→Project Symbols section, look at all
the groupings and pick out a person from the “Library\People\Animated”, and click on it to substitute it
for the triangle. Look at the picture in 3-D and enlarge it by pulling out on one of the bounding ends. Run
the simulation. You may want to change the 3-D perspective, as described previously.

Question 11: What happens to the animated people in the queues when you run the simulation?
________________________________________________________________________________

Step 2: Change what happens to the animated people when they are waiting by clicking on the
EntCustomer and expand the “Animation” details in its Properties Inspector. Change the “Default
Animation Action” to “Moving” from “MovingAndIdle” to eliminate the shaking of the animated people
when they are idle.

Step 3: Select the SrvGetIceCream SERVER object and substitute a “table” for the SERVER picture by
choosing the table under the Symbol→Project Symbols section. Repeat the procedure for the
SrvPayCashier SERVER object. Adjust the size of the SERVER objects to correspond to the persons’ size.

Step 4: Next, let’s add some stationary people behind the tables to represent the attendants. To do that,
switch to 3-D and select the “Drawing” tab. Click on the “Place Symbol” button and select a person from
the “Library\People” to represent the attendant. These non-animated library of people do not move their
body parts if used as a symbol for an entity and move by “skating” from point to point. When the person
is placed, you can click on one of the corners and hold down the Ctrl key to rotate the picture. Place the
person behind the table. Duplicate the attendant by Ctrl-C and Ctrl-V. Now do the same for the
SrvPayCashier at the cashier station as seen in Figure 2.6.

Step 5: The animated queues (i.e., the three green lines around the server objects) by default have the
entities oriented in the same direction as the entities are traveling. Modify the orientation within the queue
Introduction to SIMIO: The Ice Cream Store 27

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 as specified in Table 2.1 by selecting the Processing.Contents queue and click the Point or
Oriented Point button in the Appearances→Alignment section as seen in Figure 2.7. You can use
Inline for the InputBuffer.Contents queues.

Table 2.1: Entity Alignment Options for Animated Queues

Queue Alignment Option Description
Entities will point from the left to right no matter the
None
queue orientation within the object frame.
Inline Entities will be oriented to point forward along the line
Entities only reside on vertices of the queue but will
Point
point in the same direction for all vertices
Entities only reside on vertices but can be aligned in
Oriented Point
different directions for each vertex.15
Step 6: Repeat the process for the animated queue for the processing contents for the SrvPayCashier.
Two Vertices
Alignment Directed toward
Options Server

Figure 2.7: Changing the Orientation of Entities in the Animated Queues

Step 7: Move the symbols around on the modeling canvas to produce a better representation of the ice
cream store. You can add walls to your store by using the “Polyline” tool from the Drawing tab. Be sure
to set the height of the object in the “Object” section of the Drawing tab. You should note that the
measurements of length are in meters unless you specify the units to be different. You may need to
modify the heights of objects to make them consistent.

Step 8: The path from the SrvPayCashier object to the SnkLeave object can have a “path decorator”
added. Path decorators may be added by clicking on the path and selecting a “Decorator.” To see people
on this path, change the travel time to 300 seconds.

Step 9: Now run the simulation and look at the animation. When you run the simulation, you may need to
adjust the way it looks both in 2-D and in 3-D. The various queue symbols (lines) may need to be
extended. For instance the line associated with the number of people waiting to get ice cream may be too
short, so you might want to extend that line. Same is true for the Processing.Contents line.
Remember that the length of the animated line has no impact on the actual number in the line (i.e., there
may be more entities waiting than can be displayed but just hidden).

Question 12: Show off your animation to your friends. What is their reaction?
________________________________________________________________________________

Step 10: In the prior animation, the default entity triangle was changed to represent a person but the exact
same person arrived each time. To allow for different people to arrive, select the model entity and click the
Add Additional Symbol button under the Symbols→Additional Symbols section to create as many types of

15 Note the default queue alignment is None. Also, the Keep in Place option will force the entities to remain at the position they

enter in the model. This characteristic is seen in queues that represent systems like waiting rooms (i.e., people do not change seats).
28 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 people you would like to see arrive. Under the Active Symbol dropdown, select each symbol which will be
currently identical to the first and change the symbol following the directions in Step 1 as seen in Figure
2.8 where four total symbols have been added.16 You will need to adjust the sizes of each of the symbols.

Random
Symbol

Figure 2.8: Adding Additional Types of People

Step 11: To use the new symbols, change the Random Symbol property to True under the Animation section
of the model entity. Now, each time a person is created by the SOURCE, a random symbol will be selected
as its picture. Run the new animation and observe what happens as seen in Figure 2.6.

Step 12: Notice that the new symbols have been added to your model in the [Navigation] panel.

Part 2.4: Looking at the Results

While the animation brings attention to your model, you will be building the simulation model in order to
understand the numerical characteristics of the system being modeled. Typically, you will want to make all
the changes to the model until the model accurately represents the system and/or the potential changes have
been identified before spending a lot of time on the animation.

Step 1: First, let’s look at the results from the basic model as illustrated in Figure 2.9. Perhaps we’ll call this
the “present system” model. Under the “Run” tab, change the “Run Length” to 8 hours and run the
simulation. The results of this simulation are found under the “Results” tab. Using the “Fast-Forward”
choice will not display the animation during the simulation. Whenever the simulation is run using the
“Run” tab, it is being run in “interactive” mode allowing one to pause, step, etc. during the simulation,
even extending the run length.

Step 2: Once you run the model, you can select the Results tab to access the statistics. Notice the “Unit
Settings”. By default time is in hours, but it is easily changed using the Unit Settings button under the
Display section. The time-based statistics will display the units in parenthesis in the results.

16
The color of the clothing can be changed by selecting a color from the Decoration section and clicking on the part of the symbol you
would like colored.
Introduction to SIMIO: The Ice Cream Store 29

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 3: The results of a simulation are shown in the form a “pivot table.” A pivot table arranges the output
data according to attributes that head each column as seen in Figure 2.9.

Results Tab Filter

Output Attributes

Results Field
Re-order

Actual
Results

Figure 2.9: Looking at Results

Step 4: Each Output attribute has two symbols – one to re-order the column and the other to “filter” or
select the items displayed in the column.

Question 13: Filter the “Statistics” to show only the average values for the MODELENTITY. What is the
average Number in System and Time in System?
________________________________________________________________________________

Step 5: Attributes can be moved left and right in the display.

Step 6: Right clicking on the “Results Field” button (see Figure 2.9), you can display “show field list.”
You can drag other fields to display by dropping them in columns in the actual results.

Step 7: The “Categories” for statistics are: Content, Throughput, Capacity, FlowTime, ResourceState and
HoldingTime. These are applied to each data source as they are relevant.

Step 8: Resource utilization includes “ScheduledUtilization” as well as “UnitsAllocated,”

“UnitsScheduled” and “UnitsUtilized”. These items refer to the actual use of the capacity of the object
and the average scheduled capacity. ScheduledUtilization is computed as a ratio of the actual time the
resource is utilized divided by the total time its capacity is available. “UnitsUtilized” is the average
number of units of the resource that are utilized up to the time this report is generated.

Question 14: What is the ScheduledUtilization of the capacity of the SrvGetIceCream object?
________________________________________________________________________________

30 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 9: Notice the waiting at the server object is displayed under the “HoldingTime” of the “InputBuffer”
while the “Content” of the InputBuffer displays the number in the queue.17

Question 15: What is the average number (content) in the InputBuffer of the SrvGetIceCream server
object?
________________________________________________________________________________

Question 16: What is the average waiting time (holding time) in the InputBuffer of the SrvGetIceCream
server object?
________________________________________________________________________________

Question 17: Why does the Processing Content average number for the SrvGetIceCream server
object equal the average UnitsUtilized previously?
________________________________________________________________________________

Question 18: What is the utilization of the SrvPayCashier server object?

________________________________________________________________________________

Step 10: To switch back to the model, you need to select the “Facility” window tab, which is at the same
level as the “Results” tab.

Part 2.5: Commentary

If you have been paying close attention to the model construction, you may have noticed that:

· The SIMIO undo/redo is very convenient. This feature eases the task of trying different modeling
features and then “undoing” and “redoing” various changes. Other features of the model, such as path
type can also be easily changed and then changed back by right clicking on the link.

· You can suppress randomness in your simulation by selecting the “Advanced Options” within the Run
Setup section of the “Run” tab and “Disable Randomness”. Doing this will allow you to follow the
behavior without randomness to assist in debugging/verifying your model.

If you are familiar with Object Oriented Design (OOD), the SIMIO [Standard Library] could be appropriately
called a “class” library. The simulation objects are created from these class definitions. The act of selecting
and dropping an object onto the modeling canvas is what “instantiates” the object. Properties, as displayed in
the property inspector, define the characteristics of the objects. Later, you will see how to add your own
characteristics. Properties are initialized, but cannot be changed. Another type of characteristic, which is
called “state variable” in SIMIO, may be added if the characteristic needs to be changed during the simulation.

Generally, running the simulation with only one run (replication) does not generate reliable statistics for
results. The next chapter will demonstrate how to obtain multiple replications for a given simulation scenario.

17
The SIMIO nomenclature of “HoldingTime” and “Content” may seem non-standard. Generally, queuing theory uses the terms
“waiting time” and “number in queue”. Also “waiting line length” or “number waiting” is used sometimes instead of “number in
queue.”
Introduction to SIMIO: The Ice Cream Store 31

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Chapter 3
Modeling Distance and Examining Inputs/Outputs
For this chapter, our goal is to model people arriving to an airport and going through the check in process. In
the first phase we are only concerned with the amount time it takes the passengers to reach the security
checking station so we can determine the needed number of workers at a check-in station. Passengers arrive at
the terminal and proceed to the check-in process to get their tickets. After check-in the passengers proceed to
the security check point.

Passengers arrive according to an exponential distribution with an average of one arrival per minute.
Passengers walk to the check-in station at a rate that is uniformly distributed between two and four miles per
hour. Passengers travel 50 yards from the terminal entrance to the check in station. Following the check in
process they then must walk 65 yards to the security check point as seen in Figure 3.1. The check-in station
currently has four people assigned to process customers, who wait in a single line. The check-in process takes
between two and five minutes, uniformly distributed, to completion. The simulation model needs to be run for
24 hours.

Passengers 50 yards 65 yards Security

Check In
Arrive Check

Exponential Four Check in Stations

(1) Minute Uniform(2,5) Minutes

Figure 3.1: Airport Check-in Process

Part 3.1: Building the Model

Step 1: Create a new model, as shown in Figure 3.2. Right-click on the objects in the [Navigation] panel
and view their properties – you may need to right-click on the [Navigation] panel names and select its
properties. Call the PROJECT “AirportProblem.”

Step 2: Insert a SOURCE named SrcPassengersArrive, a SERVER named SrvCheckIn and a SINK named
SnkSecurity from the standard library to the Facility window and connect them using a path linkage (all
of this can be done using the “Add-In” section on the “Project Home” tab).18

Step 3: Click and drag an entity into the Facility window from the MODELENTITY in the [Project Library]
panel. Also change the name from DefaultEntity to EntPassenger. Change the passenger symbol to a
person from the people library (either animated or static – we will use the static people pictures
(“Woman1”) this time so you can see how they behave during the simulation).19

18
To connect objects via links without selecting them from the [Standard Library], hold the Ctrl Shift keys down while selecting the
TRANSFERNODE to start the drawing of the link. After completing the link, choose the correct link.
19
You can also add more than one symbol for the passenger as was done in Chapter 2.
32 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 3.2: The Basic Model
Step 4: Set the speed at which a passenger can walk, namely, Random.Uniform(2,4) miles per hour,
in the EntPassenger object. Set the Interarrival Time property, Random.Exponential(1) minutes,
in the SrcPassengersArrive SOURCE object. Set the Processing Time in the SrvCheckIn object to be
Random.Uniform(2,5) minutes and make sure to set its initial capacity to four.

Step 5: For the two paths, change the Drawn to Scale property of both paths to “FALSE” and set the
correct distances for each. Recall that the path from arrive to check-in is 50 yards while the path from
check-in to security is 65 yards.

Step 6: Run the simulation and observe the behavior of the “passenger” and the changing color of the
check-in.

Question 1: What does the passenger look like as they “move”?

________________________________________________________________________________

Question 2: What color is the check-in when it is idle and when it is busy?
________________________________________________________________________________

The symbol colors for the check-in can be seen by clicking on the Active Symbol (1 of 5) down-arrow in the
Additional Symbols section. It shows the five “default” states of the server.

Question 3: What is the state name and state number when the server is busy?
________________________________________________________________________________

Part 3.2: Using the 3D Warehouse

Although we have a working model, let’s change the server symbol. You can either do so by clicking the
entity or objects and changing to some of the preloaded symbols in SIMIO, or you can download new symbols
from the Trimble 3D Warehouse™ or if you have symbols20 already on your computer, you can import them
into this model.

Step 1: To employ new symbols, select the SrvCheckIn object and click one of the choices among Project
Symbols. The Trimble 3D Warehouse™ should pop up when you click the “Download Symbol” button
which will allow you to search for any symbol you want (e.g., try searching for “Airport Baggage”). Once
you find a suitable model, simply click “Download” button on the top right and SIMIO will import it for
you into a window where you can change some of its properties, such as size and orientation before
inserting the symbol into your model.

20
New symbols can be created by packages like Trimble’s SketchUp Pro™ as well as other 3-D graphics software.
Modeling Distance and Examining Inputs/Outputs 33

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 2: When you download from the 3D warehouse an “Import” window appears. First note you can
view the downloaded symbol in 2D and 3D. Also it has length, width, and height that can be specified (in
meters). A size “hint” is shown relative to a person, so you can modify size based on that. Finally you
can rotate the symbol so it has the proper “orientation”21.

Step 3: Position the baggage SrvCheckIn object so that the Input BASICNODE, the InputBuffer queue,
and the Processing queue symbols gives the appearance of a check-in at an airport as shown in Figure 3.3.

Processing (to Downloaded

accommodate 4 Symbol
persons)
Output
Input TransferNode
BasicNode

Input Buffer
Queue
Output Buffer
Queue

Figure 3.3: Baggage Check-In

Step 4: Run and adjust the animation (in 2D and 3D). Lengthen the Processing queue of the SrvCheckIn
server and modify the symbol using “point alignment” to show that up to four people can be in-process at
the check-in station at once by adding two additional vertices.

Part 3.3: Examining Model Input Parameters

Before looking more closely at the output, let’s re-examine our “input”. We can use SIMIO to understand our
assumptions about the interarrival time and processing time using input parameters. Using input parameters
instead of directly specifying the expressions allows one to use the same random expression across multiple
objects (i.e., five machines all have the same processing time) but more importantly will allow response
sensitivity and sample size error analysis to be performed.

Step 1: Click on the “Data” tab and select the “Input Parameters” icon in the View panel on the left.
Three types of input parameters can be defined: Distribution, Table Value, and Expression.22

Step 2: Click on “Distribution”, meaning we are interested in a statistical distribution. By default we are
shown a histogram sample of 10,000 observations from a “Normal distribution” whose mean is 1 and
whose standard deviation is 1.

Step 3: Let’s try some other distributions. Look at a “Triangular” with a minimum of 0.4, a mode of 0.9
and a maximum of 1.5 (that was the “get ice cream” processing time from Chapter 1).

21
Use the hint screen by toggling it by hitting the “h” key to get help on rotating, sizing and moving objects in SIMIO.
22
Distribution allows one to specify a single distribution which can be used in both sensitivity and sample size error analysis by
specifying the number of samples used to build the distribution.
34 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 4: Using the SIMIO “help”, what is the mean of a Triangular distribution whose minimum is 1,
whose mode is 3, and whose maximum is 8?
________________________________________________________________________________

Step 4: Now define the interarrival time distribution of an Exponential with mean 1 minute (Unit Type is
“Minutes”) having with the name InpDisInterarrivalTime. It’s a good practice to give each distribution
its own “Random Stream”, so we will use stream two.23

Step 5: Add a second distribution for the Check-In processing time. It is Uniform with a minimum of two
minutes and a maximum of five minutes. Name it InpDisCheckInTime and give it stream 3.

Step 6: Now go back to your model and replace: (1) the Interarrival Time in the SrcPassengersArrive to
InpDisInterarrivalTime, and (2) the Processing Time in the SrvCheckIn to InpDisCheckInTime. The
easy ways to add these input parameters is to right-click the down-arrow in the specification and select the
“Set Input Parameter”. From the options, select the name distributions.

Step 7: Run the simulation for a few minutes.

Question 5: Do you notice any change in the behavior of the simulation?

________________________________________________________________________________

Step 8: Of course you could continue to use the direct specification like Random.Exponential(1) or
you can use the “Expression” choice in the “Input Parameters” to give Random.Exponential(1) a
name to reference.

Step 9: Also you should note that a given input parameter can be used in multiple places within an object,
so changing all cases is reduced to changing the input parameter specification.

Part 3.4: Examining Output

The specific output from a simulation is a random variable produced by the entities flowing through the model
and receiving services. Their arrivals are random as well as their services. Hence one run/replication of the
model produces one experimental value. This value is not precise and to gain precision, we need to make
multiple runs/replications.

Step 1: Change the run length to 24 hours and run (i.e., use “Fast-Forward”) the simulation. After the
simulation is complete, click “Stop”, and then click the “Results” window tab to access the results
window. In the “Display” section of the SIMIO ribbon, change the units of time to minutes.

Question 6: What is the average number of people waiting at check in?

________________________________________________________________________________

Question 7: What is the average time in minutes they were waiting?

________________________________________________________________________________

Question 8: Do you think these are “good” results? Why or why not?
________________________________________________________________________________

23
A random number stream for a distribution can be viewed as a specific series of random observations from that distribution. As
such, it is reproduced by choosing that same stream again. This insures that the randomness associated with that stream is reproduced
in other scenarios, so its randomness doesn’t add additional variance to the output performance measures.
Modeling Distance and Examining Inputs/Outputs 35

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 2: To gain confidence in the precision of our results, the model needs to be run for several
independent replications. From the “Project Home” tab, let’s add a new experiment to be able to run the
model multiple times by clicking on the “New Experiment” icon.

Step 3: In the “Design” window tab, change the number of replications to ten24 and “Run” the experiment
using the icons in the “Experiment” section of the SIMIO ribbon. It should run quickly and the progress
of each replication is shown in the comment window at the bottom. If you are running with a multi-core
computer the runs will not be in order since each replication runs on a different core.

Step 4: You can look at the “Pivot Grid” or the “Reports” tab to get the results (remember that you can set
the “units” in the ribbon for the output). From the Pivot Grid, notice that the “Results Fields” now include
“Minimum”, “Maximum”, and “Half-width”. The half-width is one half of a confidence interval with the
“Confidence Level” as specified in the Experiment Properties (it is 95% by default). Recall that the
confidence interval is a statement about the precision of the summary values

Question 9: What did you get for the average, minimum, maximum, and half-width for the number of
passengers in the system?
________________________________________________________________________________

Question 10: What is the average, minimum, maximum, and half-width for time in system (FlowTime) in
minutes?
________________________________________________________________________________

Step 5: With (1 – α)% as the “confidence level,” you may recall that the corresponding statistical
confidence interval is given by the following formula.

n = number of replications
s
X ± tn -1,1-a /2 X = sample mean
n
s = sample standard deviation
tn -1,1-a /2 = critical value from t statistic
Step 6: Note that:
· The statistical summary values, such as an average, are computed average values from each
replication (one observation per replication). The confidence interval will increase in size as the
confidence level goes from 90% to 95% to 99% while the confidence interval gets smaller as the
number of replications increase.
· The “half-width” is simply one-half the width of the confidence interval
· A simulation run for n0 replications having a half-width of h0 will require (approximately) n total
replications to obtain a target half-width h according to the following equation.

h02 h = half width from observation of n replications

n @ n0 0 0
h2

Question 11: Suppose you want the average number in the system for Passengers to be within ±5% of its
mean? How many replications would be needed given the previous half width calculation?
________________________________________________________________________________

24
It should have defaulted to ten.
36 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 12: Suppose you want the maximum number in the system for Passengers to be within ±2
passengers? How many replications would be needed?
________________________________________________________________________________

Step 7: You can add the standard deviation (actually the standard deviation of the mean or standard error –
not the standard deviation of observations within a replication) by right-clicking on of the Filter Fields
and selecting “Show Field List”. Next double-click on the Std. Dev field.

Question 13: What did you get for the “Std. Dev.” for the number of passengers in the system?
________________________________________________________________________________

Question 14: What part of the confidence interval formulation has the standard error – it’s what SIMIO
calls the standard deviation here?
________________________________________________________________________________

Step 8: To remove a field, simply move it back to the PivotGrid Field List.

Part 3.5: Using Experiments

Suppose now we want to determine the effect a change in capacity of the check-in station will have? We
could simply change the capacity and re-run the simulation looking at the results. A better way is to set up a
SIMIO “experiment” which contains alternative “scenarios” that can be compared together directly.

Step 1: In the Facility view, click on the SrvCheckIn object. Right-click on the Initial Capacity property
and select “Set Referenced Property”. Use “Create New Property” and give it the name
CheckInCapacity. Doing this creates a new model “property” which is listed in the “Definitions” tab of
the model under the “Properties” listing.

Step 2: From the “Project Home” tab, create a “New Experiment” and view the “Design” Tab. Note that
this is a second designated “experiment” in the [Navigation] panel and our referenced property is now
listed in the columns.

Step 3: Add Scenario rows, changing the CheckInCapacity to two, three, and four to create an experiment
grid as seen in Figure 3.4. You can add rows by clicking on the “*” symbol in the first column.
Essentially you are defining three scenarios each having a different check-in capacity.

Reference
Property

Figure 3.4: Experiment Scenarios

Step 4: Now run the simulation performing ten replications of each scenario which is completed very
quickly.25 SIMIO will parallel process the experiments on each core of a multi-core computer.
Replications that finish turn green in parallel where the yellow scenarios are currently being run.

25
Run the experiments under the “Design” tab in the SIMIO ribbon to employ the scenario controls. If you go back to the Facility
window to run the simulation, then the controls employ their original specifications.
Modeling Distance and Examining Inputs/Outputs 37

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 5: Notice that the Pivot Grid report now includes columns for each scenario.

Question 15: What is the minimum average time in the system for “ScenarioCap4”?
_____________________________________________________________________________

Step 6: The “Reports” tab is a good way to compare scenarios. In the reports tab, the scenarios are
organized under each data item.

Question 16: What are the average and half-width for average time in system for each scenario?
_____________________________________________________________________________

Step 7: It is helpful to add “responses” to the experiment to aid in comparing scenarios. Simply click the
“Add Response” button in the “Design” tab. Insert the following two responses with expressions which
will look at the average time in system by the passengers and the total number processed specifying units
of minutes for the time in system as seen in Figure 3.5.

· TimeInSystem, in Minutes: EntPassenger.Population.TimeInSystem.Average

· TotalNumberProcessed: EntPassenger.Population.NumberDestroyed

Figure 3.5: Specifying the Time in System Response

Step 8: Reset and run the experiment again looking at the “Response Result” charts along with the scenario
grid (see Figure 3.6). To arrange the “tabs”, click the tab and drag it to the position you desire on the
display – note the positioning points. You can also right-click the tab to configure the display of results.

Step 9: The Response Results and the SMORE26 plots contain graphical displays of various summary
results from each scenario. Generally, the numerical values are hard to determine from just the graph
which is discussed later, but the graph is useful in showing trends and variation across and within
scenarios. The confidence interval on the mean, which is most often used, is displayed in the rust color.
We know statistically that if confidence intervals between scenarios don’t overlap, the scenario results are
statistically different.

26
SMORE stands for SIMIO Measure of Risk & Error.
38 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Responses

Max
Observation Response
Confidence Graph
Interval in the
rust color
Min
Observation Upper
Percentile
Mean

Lower Median
Percentile

Figure 3.6: Response Results

Step 10: Clearly, as the capacity of the check-in increases, the time in system decreases and the number
processed increases. You can modify the graph using the “View” section of the “Response Results”
options in the ribbon. The response graph displays response values (see Figure 3.6) of observations within
the simulation. The shading provides their confidence intervals which is specified in the properties of the
Experiment. Investigate clicking the “Histogram” and the “Means Line” buttons or try rotating the plot to
get a different perspective.

Step 11: If you select the “Raw Data tab found at the bottom of the “Response Results” tab, you will get the
table shown in Figure 3.7, which includes numerical details on the responses, the confidence interval
(95%) on the response, and the percentiles (which is 25% and 75% by default).27

Figure 3.7: Raw Data on Graph

Question 17: What is the mean average time in the system for “ScenarioCap3”?
________________________________________________________________________________

27
You can change the confidence level and percentiles in the EXPERIMENT properties.
Modeling Distance and Examining Inputs/Outputs 39

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 18: What is the 95% confidence interval for the mean average time in system for
“ScenarioCap3”?
________________________________________________________________________________

Part 3.6: Input Sensitivity

Since we specified the interarrival time and processing time as “input parameters” and our experiment has
responses with multiple replications greater than the number of input parameters, SIMIO can perform a
sensitivity analysis on these inputs as a by-product. Essentially, this analysis attempts to determine how the
inputs affect the responses (i.e., outputs) through simple linear regression.

Step 1: Click on the “Input Analysis” window tab for the experiment and select the “Response Sensitivity”
panel icon. Select, for example, “ScenarioCap2” and the “TimeInSystem” response.

Step 2: Four lower tabbed displays (i.e., Tornado Chart, Bar Chart, Pie Chart, and Raw Data) are available.
Each of these charts measure how the particular “input” affects the particular “response” relative to each
other. By passing the mouse over the display, the numerical values associated with the graph are
displayed. The Tornado chart shows a single response whereas the bar and pie charts show all responses
together as seen in Figure 3.8.

Figure 3.8: Tornado Chart for Sensitivity of Input to Time in System

Step 3: The tornado chart in Figure 3.8 presents the results of the regression of the two inputs,
InpDisInterarrivalTime(x1) and InpDisCheckInTime(x2), on the output TimeInSystem(y) which can be
represented more formally as y = b1 x1 + b2 x2 . By passing your cursor over the bars in the tornado chart
you can see that β1 = -291.261 and β2 = 18.993. These values have practical interpretation.
TimeInSystem(y) is negatively related to InpDisInterarrivalTime(x1) and positively related to
InpDisCheckInTime(x2). In other words, a unit increase in InpDisInterarrivalTime(x1) causes the
TimeInSystem(y) to decrease by 291.261 minutes while a unit increase in InpDisCheckInTime(x2) causes
the TimeInSystem(y) to increase by 18.993 minutes.

Step 4: In the tornado chart the solid line shows the relation between the input parameter and the response
while the outlined bar shows the negative. In this case, the lower the interarrival time (i.e., faster arrivals)
the higher the time in system. While the higher the processing time the higher the time in system. Clearly,
from the figure, the interarrival time has a much greater influence on time in system than the processing
time. However, from the bar chart the reverse is true for the total number processed. This information can
help modelers determine if more data needs to be collected to verify the input distribution assumptions.
40 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 19: Passing your cursor over the bars, what are the sensitivity coefficients for the interarrival time
and the processing time for ScenarioCap3?
________________________________________________________________________________

Question 20: When you examine the Bar Chart and Pie Chart for the time in system response, what is the
percentage contribution of each input for ScenarioCap3?
________________________________________________________________________________

Step 5: Although the impact of adding capacity to the check-in is clear looking at Figure 3.6, the capacity
can be considered an input parameter and be added to the sensitivity displays.

Part 3.7: Commentary

· The standard deviation produced in the output (see the “Results Field” option) is really the standard
error (standard deviation of the mean). No standard deviation is computed within one run of the
simulation even though it can be computed and can provide some useful modeling options (for
instance in establishing inventory policies). The danger is that the standard deviation computed
within a single replication cannot be used to compute a legitimate confidence interval, since the
observations may not be Normally distributed, nor independent and identically distributed (IID).
· The sensitivity of input parameters on individual responses opens up a useful method to explore
the behavior of the simulation model as seen in later chapters.
· Later chapters will describe in detail the SMORE plots, subset selection, ranking/selection, and
optimization.
· While other simulation languages may manage the output in a database file, SIMIO manages the
output internally but it can be exported externally.

Modeling Distance and Examining Inputs/Outputs 41

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Chapter 4
More Detailed Modeling: Airport Revisited
There is normally more than one type of check-in in an airport, as well as more than one type of passenger (i.e.
a curbside check-in or a no checked-bags check-in traveler). Our goal is to embellish the airport model from
the last chapter, to handle routing to different check-in stations, varying arrival rates of passengers, and
varying types of passengers with rate tables and data tables. We will also introduce a means of giving entities
characteristics, called state variables, which can be used to provide various distinctions.

Part 4.1: Choice of Paths

As in the previous chapter, our primary concern is with the amount time it takes the passengers to reach the
security checking station and in determining the number of needed check-in stations.

Step 1: Delete the EXPERIMENT object in the prior chapter model by right clicking on the Experiment in the
[Navigation] Panel.28 We will embellish the previous airport model to allow a choice of three different
routes for passengers arriving at the airport. Passengers who need to check bags and get tickets can either
check in at the curbside station (10%) or check in at the main station inside the airport (70%), while some
passengers can proceed directly to the security check point (20%). Refer to Figure 4.1 which depicts the
modified airline check-in process to be modeled.

One Airline Staff

(1, 2, 5) Minutes
Curbside 125 yards
Check In

10 yards 10%

Passengers 50 yards Inside 65 yards Security

Arrive 70% Check In Check

NHPP Rate Two Check in Stations

20%
Table Uniform(2,5) Minutes

125 yards

Figure 4.1: Modified Airline Check-in Process

Route1: Passengers who use the curbside check-in need to walk 10 yards to reach the station where a
single airline staff member takes between one and five minutes, with a most likely time of two
minutes, to check-in a passenger. Once they have checked in, they walk 125 yards to the security
check line.

Route2: Passengers who have checked in online and do not need to check any luggage, only need to
walk the 125 yards to the security check line once they arrive to the airport.

Route3: Passengers who plan to use the inside check-in first need to walk 50 yards to reach the station
where there are currently two airline personnel (instead of four previously) checking passengers into

28
The [Navigation Panel] in the upper right corner contains all of the models, objects, and experiments in the project. You can access
the properties, rename the item, or delete the item from the project.
42 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 the airport. It takes between three and ten minutes uniformly to process each passenger. Once they
have checked in, they then walk 65 yards to reach the security check line.

Step 2: Modify the model from last chapter to include the new curbside check-in (SERVER) as well as the
direct path (i.e., Route 2) from the arrivals to the security gate as seen in Figure 4.2. Specify the properties
according to description in Figure 4.1.

Route1

Route2

Route3

Figure 4.2: Creating Different Routing Paths

Step 3: It has been observed that passengers follow the probabilities in Table 4.1 in determining which
route they will follow.

Table 4.1: Passenger Type Percentages

Check-in Type Percentage
Curbside Check-in 10%
Inside Check-in 70%
No Check-in (Carry on) 20%

Step 4: To model the decision of a passenger to choose their check-in route the Selection Weight property
of Path objects will be utilized. The probability of choosing a particular path is that path's individual
selection weight divided by the sum of all path selection weights from that node object. Place the proper
weights to each different check-in route (see Figure 4.3).

More Detailed Modeling: Airport Revisited 43

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Selection Weight

Figure 4.3: Weighting Paths

Step 5: Remember to select the Outbound Link Rule to “By Link Weight” on the TRANSFERNODE
Output@SrcPassengersArrive (see Figure 4.4).29

Link Rule

Figure 4.4: Outbound Link Rule

Step 6: Run the model for one, twenty-four hour replication (no “Experiment” needed). You might want to
improve on our animation.

Question 1: What is the average number of passengers waiting at Curbside and the Inside Check-in?
________________________________________________________________________________

Question 2: What is the average wait time at each Check-in?

________________________________________________________________________________

Question 3: What is the average time it takes for a passenger to check-in (from entry to security)?
________________________________________________________________________________

Part 4.2: Changing Arrival Rate

Passengers do not arrive at a constant rate (i.e., homogeneous process) throughout the entire day in a real
airport (even though they may arrive randomly), so we should expand our model to handle such phenomena.
Our approach will be to allow the arrival rates to change throughout the day. To model this changing rate, we
will use a data object in SIMIO called a Rate Table and reference the Rate Table we create in the arrival logic
of our Source object. It is important to understand that the interarrival times for this rate will be assumed to be
Exponentially distributed, but the mean of this distribution changes with time. If an interarrival process has an

29
Although “Shortest Path” is the default Outbound Link Rule, “By Link Weight” rule will still be used unless the entity destination
has been assigned to a particular node. Therefore, we could have used the default rule.
44 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Exponential distribution, then it is known statistically that the number of arrivals per unit time is Poisson
distributed. Since the parameter (its mean) changes with time, it is formally referred to as a “non- “non
homogeneous Poisson” arrival process, also called a NHPP (Non-Homogeneous Poisson Process). Therefore
when you use a SIMIO Rate Table you are assuming the arrival process is a NHPP NHPP. Table 4.2 shows the
hourly rate over the 24 hour day where more passengers on average arrive during the morning and dinner time
hours. The zero rates imply no arrivals during that time period.

Table 4.2: Hourly Arrival Rate During Each Hour

Begin Time End Time Hourly Rate
Midnight 1am 0
1am 2am 0
2am 3am 0
3am 4am 0
4am 5am 0
5am 6am 30
6am 7am 90
7am 8am 100
8am 9am 75
9am 10am 60
10am 11am 60
11am Noon 30
Noon 1pm 30
1pm 2pm 30
2pm 3pm 60
3pm 4pm 60
4pm 5pm 75
5pm 6pm 100
6pm 7pm 90
7pm 8pm 30
8pm 9pm 0
9pm 10pm 0
10pm 11pm 0
11pm Midnight 0
Step 1: Arrivals are only present from 5am to 8pm for a total of 15 hours each day. A simulation run
length of 24 hours would cause the time-based statistics like number in station, number in queue, and
scheduled utilization to be computed when there were no arrivals, causing lower than expected values for
these statistics. Therefore we will set the Starting Time within the “Run Setup” tab to 5:00 am and the
Ending Type to Specific Ending Time at 8:00 pm as shown in Figure 4.5

Figure 4.5: Replication Start and Stop Time

Step 2: To create a RATE TABLE, select the “Data” tab and then click Create →Rate Table. Name the table
PassengerArrivalRate. Set the Interval Size to one hour and the Number of Intervals to 15, corresponding
to the run length as seen in Figure 4.6. While you can modify the size (e.g., half hour increments) and
number of time intervals during which the arrival rate is constant, you cannot modify the rate, which is
fixed at arrivals per hour.

Step 3: Enter values in the rate table as shown in Figure 4.6.

More Detailed Modeling: Airport Revisited 45

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 4.6: The Rate Table (15 intervals)
Step 4: Modify the SrcPassengersArrive source accordingly (see Figure 4.7 for details). SIMIO refers to
the NHPP as a “Time Varying Arrival Rate” ARRIVAL MODE.

Modify Arrivals

Figure 4.7: Specifying a Time Varying Arrival Rate

Step 5: Run the model again and observe the results.

Question 4: What are the average Wait Times at each check-in location (Curbside and Inside)?
________________________________________________________________________________

Question 5: What is the average check-in time for a passenger (Curbside and Inside)?
________________________________________________________________________________

If the length of the simulation (replication/run length) exceeds the length of the rate table, then the rate table
simply repeats until the simulation replication is terminated. So if the run length should be longer and there
should be no arrivals during that time, then the rate table should contain zeroes accordingly.

Part 4.3: State Variables, Properties, and Data Tables

Properties and state variables can be characteristics of the “Model” or of the “ModelEntities”. A “property” of
an object is a part of the object’s definition and is assigned at the time an object is created, but it can’t be
changed during the simulation. Properties can be expressions, so their evaluation and the time of their creation
can yield different values. In contrast, a “state variable” is a characteristic of an object that can be changed
46 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 during the simulation. Generally, when properties and state variables are associated with the model, they are
global in the sense that their values are visible throughout the model. When properties and state variable are
associated with entities, then you need to have access to that model entity in order to access the state variable
value.

In an actual airport more than one type of passenger arrives. Some are continental travelers while others are
inter-continental travelers. Still there are some passengers which have handicaps and require special attention
or are with families. The different types of passengers result in different processing times. To model these
three cases we will use a combination of state variables and properties to store the information and to reference
the information when needed.

There is a 33% chance for each type of passenger to arrive to our airport. The processing times at the curbside
check-in and the inside check-in varies with the type of passenger. Specifically, see Figure 4.8 for the
processing times.

Passenger Curbside Check-In Inside Check-In Time

Type Time (minutes) (minutes)
1 Triangular(1, 2, 5) Uniform(2, 5)
2 Triangular(2, 3, 6) Uniform(3, 5)
3 Triangular(3, 4, 7) Uniform(4, 6)
Figure 4.8: Check-In Times for Different Passenger Types
The MODELENTITY will need to reference these data as specifications at the check-in counters

Step 1: Since the passenger type is not a general characteristic of SIMIO model entities, we need either to
find a SIMIO characteristic we can use to represent the type or define such as characteristic ourselves.
User-defined characteristics whose values can be re-assigned are the “STATE VARIABLES.” A state
variable can be a characteristic of the “Model” (namely global) or a characteristic of the “ModelEntity”.
Here we are interested in a characteristic of a MODELENTITY.

Step 2: To create a MODELENTITY State Variable, first select the “ModelEntity” object in the Navigation
panel. Click “States” panel icon in the “Definitions” tab. It is convenient to number our passenger type,
so select a DISCRETE “Integer” state from the ribbon. Let’s name it EStaPassengerType30. Next we need
to assign each MODELENTITY a value for passenger type.

Question 6: When you define your new state variable, what other state variables are already defined for
MODELENTITIES?
________________________________________________________________________________

Step 3: We need to assign a value to EStaPassengerType for each entity. To obtain a random assignment
of passenger type to entities (i.e., 33% chance for each type), we must use the Discrete Distribution
(entered in cumulative distribution form).

Random.Discrete(1, 0.33, 2, 0.66, 3, 1.00)

Step 4: One convenient place is to make this assignment at the SrcPassengersArrive source object in the
“State Assignments” property. Click on the plus box to reveal “Before Exiting”. Then click the ellipses
box at the end of that specification to bring up the “REPEATING PROPERTY EDITOR”31, which in this case
allows state value assignments.

30 We will use “E” to when denoting “Entity” and “Sta” to denote “State variable”.
31
It’s called a Repeating Property Editor, even though, in this case, it is assigning state variables
More Detailed Modeling: Airport Revisited 47

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Repeating Property
Editor

Discrete Distribution where EStaPassengerTyp

each passenger type is e
equally likely (one-third) State Variable

Figure 4.9: Assigning Passenger Type

Now each passenger has a (randomly assigned) state variable that describes its passenger type.

Step 5: Next we need to store the information for the check-in times. The expression of these times (e.g.
Random.Unform2,5)) do not change throughout the simulation. And these check-in times need to be
accessed for all entities. The check-in times are not specific to each entity, but are characteristics of the
model. So, “properties” are an appropriate way to store these data for the model.

Step 6: A way to organize model properties is through the use of a DATA TABLE. The data table is
functionally identical to the table we created for Figure 4.8. To create a DATA TABLE click “Tables” in the
“Data” tab, then “Add Data Table.” Call it TablePassenger. Characteristics, called “properties”, can be
added to the data tables – these properties are the columns in the table. In our DATA TABLE, the processing
times for each of the three types of passengers for each check-in location will be specified. Select the
properties from the “Standard Property” option in the Properties section. PassengerPriority will be an
Integer property whereas the check-in times will be Expression properties named CurbsideCheckInTime
and InsideCheckInTime.32 Priorities 1, 2, and 3 represent Continental, Inter-Continental, and special
needs passengers respectively as in Figure 4.10.33

Properties

Figure 4.10: The Data Table

Step 7: Since the check-in times are in minutes, we want to specify the “Unit Type” for these properties as
seen in Figure 4.1134.

32Note that if you first specify the “General” name, it will become the default for the “DisplayName” property. Also, by using proper
case, the table labels as seen in the Figure 4.10 will have spaces between the words.
33Even though the check-in times are expression properties, the familiar drop down expression editor is not available in DATA TABLES.
Therefore, the easiest way is to put in all three priorities and type in the first value of the first row. Then copy it to the other rows and
just modifying the values of the parameters.
34
It may seem obvious the Unit Type for the time properties should be specified. However, one can leave the units as unspecified then
the numbers will take on the units of the property when they are used (i.e., default value of hours).
48 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Specify Unit
Type Time
(Minutes)

Figure 4.11: Inside Check-In Properties

Step 8: Now we must specify the processing times at the check-in stations to depend on the particular
entity’s passenger type, shown in Figure 4.12. The passenger type is used as the “row reference” in the
table. Thus, the specifications needed for the new processing times use the EStaPassengerType state
variable as:

· At SrvCurbSideCheckIn:35
TablePassenger[ModelEntity.EStaPassengerType].CurbSideCheckInTime
At SrvCheckIn:
TablePassenger[ModelEntity.EStaPassengerType].InsideCheckInTime
Notice that the content of the []36 are used to designate the specific row in the table, similar to arrays in
programming languages. Unfortunately, the expression editor does not know about the DATA TABLE
properties since they are user-defined, so you must type in the property into the expression. Also be sure that
the processing time units in the check-in stations are specified in minutes.

Figure 4.12: Utilizing the Table to Specify the Processing Times

Step 9: Run the model. Remember no arrival occurs before 5am and the simulation is started at 5am so no
animation appears before then.

35To enter this specification so as to employ the expression editor during entry is to first put in the table name and property as
TablePassenger.CurbSideCheckInTime and then insert [ModelEntity.EstaPassengerType].
36 Note, table rows start at one (i.e., the first check in time is TablePassenger[1].InsideCheckInTime)
More Detailed Modeling: Airport Revisited 49

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 7: What is the average wait times at each check-in location (Curbside and Inside)?
________________________________________________________________________________

Question 8: What is the average check-in time for a passenger?

________________________________________________________________________________

Properties and State Variables: Before leaving this section, let’s reflect on the characteristics “properties”
ey will be in frequent use. Properties and state variables can be characteristics
and “state variables” because they
of the MODEL or of the MODELENTITIES. A “property” of an object is a part of the object’s definition and is
assigned at the time an object is created, but it can’t be changed during the simulation. Properties can be
expressions, so their evaluation can yield different values. In contrast a “state variable” is a characteristic of
an object that can be changed during the simulation. Generally, when properties and state variables are
associated with the model, they are global in the sense their values can be obtained almost anytime. When
properties and state variables are associated with entities, then you need to have access to that model entity in
order to access the value.

SIMIO makes extensive use of properties and state variables, which you can see by looking at their
“definitions” and expanding the (Inherited) Properties/States. Some properties and states are related. For
example, the “InitialPriority” property of a MODELENTITY becomes the initial value of the “Priority” state
variable. Other states are given their initial values by properties.

Part 4.4: More on Branching

At times, the waiting lines at the inside Check-in station seem overcrowded. We decided to add a second
Inside Check-in station that passengers may use during rush hours. Thus, passengers will be routed to
whichever Check-in station has the smallest number in queue. The distance from this second Check-in station
to the Security Check is 75 yards as seen in Figure 4.13.

One Airline Staff

(1, 2, 5) Minutes
Curbside 125 yards
Check In

10 yards 10%
Inside 65 yards Security
Check In Check
Passengers 50 yards
TransferNode
Arrive 70%
Inside 75 yards
NHPP Rate Check In 2
Table 20%

125 yards

Figure 4.13: Airline Check-in Process with Two Inside Check In Stations
Step 1: Add the second check-in station. Call it SrvCheckIn2. Assume it has a capacity of one and it’s
processing times are the same as those for the regular check-in station. Add the path of 75 yards from this
second station to SnkSecurity.

Step 2: We need to model the choice made by the “inside” check-in passengers to go to the check-in
station with the smallest number in queue. There are several ways to model this situation. Perhaps the
most obvious way is to add a TRANSFERNODE that branches passengers to the shortest queue. The distance
from the SrcPassengersArrive to this TRANSFERNODE is 50 yards (remember that the selection weight is
70 for this path and you will have to delete the original path to the SrvCheckIn object).

50 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 · The TRANSFERNODE should specify the Outbound Link Rule as “By Link Weight”
· Set the Selection Weight properties of the two paths from the TRANSFERNODE (the “value” of a
“true” expression is one and zero if it is “false”)

(1) TO SrvCheckIn:
SrvCheckIn.InputBuffer.Contents.NumberWaiting <=
SrvCheckIn2.InputBuffer.Contents.NumberWaiting

(2) TO SrvCheckIn2:
SrvCheckIn2.InputBuffer.Contents.NumberWaiting <
SrvCheckIn.InputBuffer.Contents.NumberWaiting

Question 9: Which check-in station is chosen if there are “ties” in the size of the InputBuffer (For
example, suppose both are empty)?
________________________________________________________________________________

Question 10: What is the utilization of SrvCheckIn2 and does it make sense to have both check-ins?
________________________________________________________________________________

Part 4.5: Work Schedules

Suppose you decide only to staff the second inside check-in station from 10am to 6pm. This kind of concern
requires a work schedule which alters the capacity of a server over time. Work schedules will be built from the
“Schedules” component in the “Data” tab.

Step 1: Click on “Work Schedule” button under in the “Schedules” icon in the “Data” tab to add a new
work schedule. By default, the StandardWeek schedule has already been defined as seen in

Step 2: Figure 4.14. Rename this schedule SecondCheckInStation. Change the Days to “1” – this
schedule will repeat daily.37 For each day of the work schedule, you need to specify the work pattern for
that day. Therefore, you can have different patterns for different days (i.e., weekends, etc.). The
StandardDay pattern was specified for Day 1. The Start Date property can be left as the default since we
have a daily pattern.

Your “default” date will be different

Figure 4.14: Creating SecondCheckInStation WORKSCHEDULE

Step 3: In the “Day Patterns” tab, you can add DAY PATTERNS by clicking the Day Pattern button in the
“Create” section. SIMIO defines the StandardDay pattern of 8am to 5pm with an hour lunch at 12pm.
Since we start the day at 10am and end the day at 6pm, we only need one work period in the pattern.
Select the second work period (1pm to 5pm) and delete it. Then specify the Start Time to be 10am and the
Duration to be eight hours or the “End Time” to be 6pm as seen in Figure 4.15. Specify a “1” for the

37
If there are seven days in the pattern, then the days of the week will appear. In our case there is only the Day 1 pattern. SIMIO will
repeat the pattern based on the number of days in the WORK SCHEDULE. For each day in the work schedule you need to specify a pattern.
Also, we have the ability to specify exceptions to the day pattern or work periods for particular dates.
More Detailed Modeling: Airport Revisited 51

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Value which will specify the capacity of the second inside check in to be one. If a work period has not
been defined for a period in the day pattern, it is assumed to be off shift (i.e., zero capacity).

Figure 4.15: Setting a Day Pattern with Different Work Periods

Step 4: Go to the “Facility” view and click on the SrvCheckin2 object. Change the Capacity Type to
“WorkSchedule” and the Work Schedule property to SecondCheckInStation as in Figure 4.16.

Work
Schedule

Figure 4.16: Specifying a Work Schedule

Step 5: Now we need to be sure that people who choose the inside check-in only go to the SrvCheckIn2
station if it is open (i.e., only between hours 10 and 18). This requires another TRANSFERNODE as seen in
Figure 4.17. Connect the original TRANSFERNODE to the new one via a Connector and connect the new
TRANSFERNODE to the inputs of both check-in stations.

Input@SrvCheckI
From n
SrcPassengersArrive
TRANSFERNODE to
choose shortest Queue
New TRANSFERNODE to
choose open
Input@SrvCheckIn
SrvCheckIn
2

Figure 4.17: Adding Another TransferNode to Handle the Logic

· The new TRANSFERNODE should determine routing By Link Weight as well.
· Path to Input@SrvCheckIn has zero length38 and Selection Weight property:
Run.TimeNow <=5 || Run.TimeNow>=1339
· Path to TRANSFERNODE to choose shortest queue has zero length and Selection Weight property of
Run.TimeNow > 5 && Run.TimeNow <1340

38
May be easier to use a CONNECTOR instead of a PATH – CONNECTORS have zero length and take zero time to traverse.
39
Note, SIMIO utilizes || as the “Or” logical operator and && as the “And” logical operator. Also, this expression works only here
since we are simulating just one day and 10 A.M. represents five hours into the simulation since we are starting at 5 A.M. If we were
simulating more than one day, then we would need to specify Math.Remainder(Run.TimeNow, 24)*24 which converts the
running time into a number between 0 and 24.
52 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 6: Note that when there is no context, the default units of time are hours. Now run the model and
obtain these basic statistics.

Question 11: What is the average Check-In time for a Passenger?

________________________________________________________________________________

Question 12: What is the average wait times for Curbside Check-In and Inside Check-In?
________________________________________________________________________________

Question 13: What is the utilization for an employee at the Curbside Check-in?
________________________________________________________________________________

Question 14: What are the utilizations for an employee at the Inside Check-in?
________________________________________________________________________________

Step 7: You may have noticed that in the simulation, there are entities in the queue of the SrvCheckIn2
that are in line at the time the station closes. These are marooned in this model. We will consider how to
handle this kind of problem in a later chapter.

Question 15: What would have to change to the model every time the work schedule of the second check in
server was modified?
________________________________________________________________________________

Step 8: If we wanted to determine the best starting time and ending time to have the second check in
station open, the link weights using the Run.TimeNow expression would have to be changed to match
the new start and end times each time. A better approach would be to use the capacity of the server so as it
goes to zero we need to close the path and open the path when the capacity goes to one. Change the link
weights according and rerun the model comparing the previous results.

· Path to Input@SrvCheckIn Selection Weight property: SrvCheckIn2.Capacity == 041

· Path to TRANSFERNODE to choose shortest queue has zero length and Selection Weight property of
SrvCheckIn2.Capacity > 0

Part 4.6: Commentary

· The specification of the time-varying arrival pattern in SIMIO requires fixed units for the arrival rate,
namely arrivals per hour.
· Pay special attention to the SIMIO specification of cumulative probability and its value for the
Random.Discrete and Random.Continuous distributions.
· While work schedules in SIMIO offers considerable flexibility in terms of specifying exceptions etc.,
they are somewhat complicated to input. One has to be careful in specifying the correct times and
duration. Also be careful in specifying the time to begin and end the simulation run.
· In SIMIO, when a server’s capacity goes to zero while there is an entity (or entities) being processed,
SIMIO chooses, by default, to “Ignore” this so that the items “in process” will be processed while at
the same time the server capacity goes to zero. If another behavior is desired, then you will need to
implement it using more advanced concepts, which will be described later.

40
We could have modified the original link weights and avoided having to add the new paths and transfer node.
Specify the expression to be (TimeNow <=5 || TimeNow >= 13) || SrvCheckIn.InputBuffer.Contents.NumberWaiting <=
SrvCheckIn2.InputBuffer.Contents.NumberWaiting for the link weight for the path to SrvCheckIn and for the path link
weight to SrvCheckin2 to (Run.TimeNow >5 && Run.TimeNow < 13) && SrvCheckIn2.InputBuffer.Contents.NumberWaiting
< SrvCheckIn.InputBuffer.Contents.NumberWaiting.
41
Note, SIMIO utilizes a double equal sign (i.e., “==”) as the logical equal.
More Detailed Modeling: Airport Revisited 53

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 · Note that if the length of the simulation replication exceeds the work schedule time, then the work
schedule is repeated until the replication ends.

54 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Chapter 5
Data-Based Modeling: Manufacturing Cell
Most simulation models employ substantial amounts of data in their development and execution. In SIMIO,
employing a data table, as seen in the previous chapter, provides a convenient means to store and recall the
data. In this chapter we will use tables more extensively by re-constructing a manufacturing cell consisting of
several work stations.

A small manufacturing cell consists of three workstations through which four part types are processed. The
workstations, A, B, and C, are arranged such that the parts must follow the one-way circular pattern as shown
in the layout in Figure 5.1. For instance, if a part is finished at Station C, it would have to travel the distance to
Station A, then the distance to Station B before it could travel to where it exits the system. We want to use a
simulation of a five day week with eight hours of operation a day for a total of 40 hours.

One Machine Two Machines

(2, 5, 8) Minutes (5, 9,11) Minutes

Station A Station B

Parts have different 25 yds 35 yds 20 yds

part routings
Parts Travel Direction Parts
Arrive Leave
40 yds 30 yds
Parts move at 2 mph

Station C

Two Machines First 4 hours

One Machine Second 4 hours
(2, 6, 11) Minutes

Figure 5.1: Workstation Layout

All parts arrive at the “Parts Arrive” location and leave at the “Parts Leave” location. Travel between
workstations is a constant two miles per hour. Note SIMIO’s default speed for entities is meters per second.
The distances (yards) between workstations are shown in Table 5.1 and in Figure 5.1.

Table 5.1: Distance between Workstations

WorkStation Path Distances (yds.)
Parts Arrive to Station A 25
Station A to Station B 35
Station B to Parts Leave 20
Parts Leave to Station C 30
Station C to Parts Arrive 40

Each part type arrives as follows.42

· Part 1’s arrive randomly with an average of 15 minutes between arrivals.
· Part 2’s have an interarrival time that averages17 minutes with a standard deviation of 3 minutes.

42
For the moment, we will let you think about the distributions that these parameters may represent.
Data-Based Modeling: Manufacturing Cell 55

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 · Part 3’s can arrive with an interarrival time of anywhere from 14 to 18 minutes apart.
· Part 4’s arrive in batches of five exactly 1 hour and 10 minutes apart.
Question 1: What distributions will you choose to model these four arrival processes?
________________________________________________________________________________

Each of the part (entity) types is sequenced (routed) across the stations differently. In fact, not all part types
are routed through all machines, as seen in the sequence order given in Table 5.2.
Table 5.2: Part Sequences
Step 1 2 3
Part 1 Station A Station C
Part 2 Station A Station B Station C
Part 3 Station A Station B
Part 4 Station B Station C
Server Properties:
· Station A has one machine with a processing time ranging from two to eight minutes and a likely
(modal) time of five minutes.
· Station B has two machines, each with a processing time ranging between five and eleven minutes and
a likely time of nine minutes.
· Station C has two machines that operate during the first four hours of each day and one machine
operating during the last four hours of the day. Each has a processing time ranging between two and
eleven minutes and a mode of six minutes.

Question 2: What distributions will you choose to model the three station processing times?
________________________________________________________________________________

Part 5.1: Constructing the Model

We will build the SIMIO model using four SOURCES, three SERVERS and several types of links.

Step 1: Insert four SOURCES named SrcParts1, SrcParts2, SrcParts3, and SrcPart4 to create the four
different part types (i.e., drag four MODELENTITIES onto the model canvas named EntPart1, EntPart2,
EntPart3, and EntPart4).43 Four SOURCES are necessary because each part has its own arrival process.
Make sure to change the Entity Type property of the four SOURCES to produce the appropriate entity type.

Step 2: Insert three SERVERS (i.e., one for each workstation named SrvStationA, SrvStationB and
SrvStationC) and one SINK named SnkPartsLeave. Be sure to name all the objects, including the
entities, so it will be easy to recognize them later as seen in Figure 5.2.

Step 3: Probably the best method to simulate the circular travel pattern would be to use a BASICNODE at
the entry and exit point at each station.44 This allows the entities to follow the sequence pattern without
having to go through an unneeded process. A BASICNODE is a simple node to connect two different links
while a TRANSFERNODE can connect different links as well but provides the ability to select destination,
path, and/or require a TRANSPORTER.

43
To do this efficiently, drag one SOURCE on to the facility and name it SrcParts. Then copy and paste it three times which will take
advantage of the SIMIO naming scheme (i.e., it will create SrcParts1, SrcParts2, and SrcParts3). Now, rename the first source
SrcParts4. Repeat this method for the four entities starting with EntPart.
44
A TRANFERNODE used to enter/exit from a station can disrupt the SIMIO internal sequencing of sequence steps if one is not careful.
56 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 4: Now connect all the objects using CONNECTORS and PATHS. Note that because there is no time lapse
or distance between the circular track and the server objects, we will use connectors. The distances
between stations will be calculated using the five paths connecting each entry and exit node. Be sure that
you draw the connectors and paths in the right direction – zoom in on your model to see the “arrows” as
shown in Figure 5.2 (i.e., links are entering the input nodes of the SERVERS while leaving the output
nodes).

Basic Node

Paths

Connector
s

Figure 5.2: Initial Model

Step 5: Change the properties of each entity so that they now travel at the desired speed of two miles per
hour.45 Also change the color of the entities so you can distinguish between the four part types.

Step 6: Change the properties of the four source objects so that they correspond to the correct parts and
have the desired interarrival times and entities per arrival. We chose an Exponential, Normal, and
Uniform as the distributions of interarrival times for part types one, two, and three respectively with the
parameters given. For part type four, we chose a constant interarrival time of 70 minutes and Entities Per
Arrival property set to 5 as seen in Figure 5.3.

Figure 5.3: Specifying Parameters for the SrcPart4

Step 7: Set the distances for each of the five paths so they correspond to the distances between the stations.
You will have to set the Drawn To Scale property for each to “False” in order to do this and then specify
the logical length and units of type “Yards”.

45
Select all four entities using the Ctrl button and specify the speed once for all entities.
Data-Based Modeling: Manufacturing Cell 57

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 8: Set the processing times for all the stations using the Pert (BetaPert) distribution. In general, we
prefer the Pert distribution to the Triangular distribution. It is specified like the Triangular
(minimum, most likely, maximum), but has “thinner” tails, which, we believe, better represents the
expected behavior of the processing time distribution. See Appendix A for more information.

Part 5.2: Setting Capacities

We will need to set the capacities for the three workstations A, B, and C.

Step 1: Set the capacities for A and B to one and two respectively. For Station C, we will need to add a
work schedule to accommodate the changes in capacity.

Step 2: Go to the “Data” tab and select “Schedules” and modify the default schedule (StandardWeek) and
the default day pattern (StandardDay). Rename the work schedule to ScheduleForC46 and leave the Start
Date property alone to remain the default date the simulation will start in the Run setup.

SIMIO will repeat the work cycle over how many days are in the work pattern. Because our simulation
will start at 12am and run for 40 hours we want to set the Days in Work Schedule property to 1 and let the
cycle repeat after the first 24 hours.

Figure 5.4: Setting up the WorkSchedule

Now select the “Day Patterns” tab to modify the StandardDay day pattern to match the one in Figure 5.5.
Change the Start Time property of the first row to 12:00 AM which will cause the Duration to change to
match the End Time. Change it back to “4 hours” and click the Enter. After changing the value of the
capacity to two, repeat this process for the second row using one as the capacity value. Next, add four new
work periods by specifying the Start Time, the Duration, and the capacity value.

Figure 5.5: Modifying the Day Pattern to Represent Three Eight Hour Days
Step 3: Go back to the “Facility” tab and set the Capacity Type property of Station C to “WorkSchedule”
and add your work schedule, namely ScheduleForC.

46
Double clicking the Standard Week name will allow you to change the name.
58 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 5.6: Specifying the WorkSchedule for the StationC
Question 3: For how many hours during a 24 hour day is the capacity of Station C equal to two?
________________________________________________________________________________

Step 4: Save and run the model for 40 hours observing the path the different entities travel.

Question 4: How long is each part type in the system (enter to leave)?
________________________________________________________________________________

Question 5: Why do you think the time in the system is large for all four part types?
________________________________________________________________________________

Part 5.3: Incorporating Sequences

The way the current model is set up, entities will branch at nodes based on link weights and the direction on
the link. However, we want the entities to follow a desired sequence of stations as seen in Table 5.2. We
would like to be sure that entities will continue along the circular pattern until they reach the node that takes
them to their particular station. To accomplish this we will use “Sequence Tables” (from the “Data” tab).

Step 1: Click the “Add Sequence Table” button in the tab and rename the new table SequencePart1 to
correspond to the first entity type.

Step 2: Now add the sequence of nodes that you want Part 1 to follow (i.e., Station A, Station C and Exit).
You only need to include the visited nodes at each station as well as the exit. Once the entities sequence
has been set, it will always travel the shortest route to get to the first node on the list. The correct sequence
for Part1 is given in Figure 5.7.47

Be sure to include
the exit in the
sequence

Figure 5.7: Sequence for Part1

Step 3: Next, add three more sequence tables to correspond to the remaining three entities (i.e.,
SequencePart2, SequencePart3, and SequencePart4). Set the sequences for each of the other three parts
in their corresponding sequence tables as well. Make sure the last node is associated with the SINK. Note
that the station’s names may appear differently in your sequence if you named them differently.

47 Note the input nodes of the associated SERVERS and SINK are specified rather than then actual object. Note the names of the nodes are
in the form Input@ObjectName.
Data-Based Modeling: Manufacturing Cell 59

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 4: Now go back to the “Facility” tab and click on the EntPart1 entity. Set the Initial Sequence
property to the “SequencePart1” under the “Routing Logic” section which you just created for “Part 1” as
seen in Figure 5.7. Do the same for each of the other entity types and their associated sequences.

Figure 5.8: Specifying an Entity to follow a Particular Sequence

Step 5: For each of the seven TRANSFERNODES (output nodes at each source and server), change the Entity
Destination Type property to “By Sequence” for each one.48 The first time an entity passes through a
transfer node where its destination is “By Sequence”, the node sets the destination of the entity to the first
row index of the SEQUENCE TABLE. After that, every time an entity passes through a transfer node with
the “By Sequence” destination type the row index of the table is increased by one and the destination of
the entity is set to the next node in the table.49

Table 5.3: Definition of Basic and TransferNodes

Simple node to support connection between links. Often used for Input nodes of
objects (e.g., SINK, SERVER, etc.).
Supports connection between links but has the ability to select entity destination as
well as request a transporter. Output nodes of objects (SOURCE, SERVER, etc.) are
TRANSFERNODES

Question 6: Will passage through a BASICNODE also change the row index of the SEQUENCE TABLE?
________________________________________________________________________________

Question 7: Will passage through a TRANSFERNODE always change the row index of the SEQUENCE TABLE?
________________________________________________________________________________

Step 6: It will also help to make sure they follow the correct sequence. Run the simulation for 40 hours at a
slow speed to be sure each of the parts follows their sequence appropriately.

Question 8: Turn off the arrivals of all but one part type by changing the Entities Per Arrival property of the
SOURCE object to zero for all but one SOURCE. Does it follow the expected sequence?
________________________________________________________________________________

Step 7: If you turned off any source, turn them back on by changing the Entities Per Arrival property back
to the original value. Save the current model and run it for 40 hours.

Question 9: How long is each part type in the system now (enter to leave)?
________________________________________________________________________________

48 To set them all at once, press the Ctrl key while selecting all seven TRANSFERNODES and then set the Entity Destination to “By

Sequence.”
49If the current row index is on the last row and the entity enters a TRANSFERNODE that has “By Sequence” before reaching the current
destination node, the row index will start over at one.
60 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Part 5.4: Embellishment: New Arrival Pattern and Processing Times

We now learn that the parts arrive in one stream of arrivals (i.e., with a mean arrival rate of 6 per hour, Poisson
distributed50), instead of four. There are still four part types but the type is random, based on the percentages
specified in in Table 5.4. The number of each that arrive doesn’t change (meaning Part 4 arrives 5 at a time).

Table 5.4: Part Type Percentage

Part Type Percentages Number to Arrive
Part 1 25% 1
Part 2 35% 1
Part 3 25% 1
Part 4 15% 5

Question 10: When do you think a single source type of model is preferred over a multiple source model?
________________________________________________________________________________

Previously, the processing times at the various stations were the same regardless of the part type. Now we
discover the processing times depend on the particular part type as given in Table 5.5. In this table, the
sequences are the same as before, but here the processing times have been added.

Table 5.5: Entity Sequences and Processing Times (in Minutes)

Step 1 2 3
Part 1 Station A (Pert(2,5,8)) Station C (Pert(2,6,11))
Part 2 Station A (Pert(1,3,4)) Station B (Uniform(5 to 11) Station C (Uniform(2 to 11))
Part 3 Station A (Triangular(2,5,8)) Station B (Triangular(5,9,11))
Part 4 Station B (Pert(5,9,11)) Station C (Triangular (2,6,11))

Question 11: Do you think that it is realistic that the processing time for Part 1 on Station A might be
different for Part2 on that same Station A?
________________________________________________________________________________

Step 1: Now delete all but one of the SOURCES referring to Figure 5.9 as an example renaming the one
SOURCE as SrcParts. We will return to specifying this source after we specify the processing times.

50
Remember that this Poisson arrival rate is equivalent to an Exponential interarrival time with a mean of 10 minutes. SIMIO wants
the interarrival time, not the arrival rate – so be prepared to make the conversion.
Data-Based Modeling: Manufacturing Cell 61

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Single
Source

Figure 5.9: Manufacturing Cell Model

Step 2: To model the processing times which are dependent on the part type, the processing times will be
added to the original sequencing tables. When the entity is assigned a destination (i.e., a row in the
sequence table is selected), the processing time column can be accessed for that a row. (Note, you could
use a more complicated expression for the processing time by checking to see if this is a Part A, then make
the processing time X, etc. but is not very easy if there is a large number of parts).

Step 3: Go to the “Data” tab and click on the SequencePart1 table which is associated with Part 1 types.
Since the processing times can be any distribution (i.e., any valid expression), add a new Expression type
column from Standard Property drop down named ProcessingTimes. Make sure the column has the
appropriate units by making the Unit Type be “Time” and the Default Units to be “Minutes.”

Step 4: Set the Processing times for SrvStationA (i.e.., the first row) to a Random.Pert(2,5,8),
SrvStationC (i.e., the second row) to a Random.Pert(2,6,11) and finally the SnkPartsLeave or
third row will remain zero as seen in Figure 5.10.

Figure 5.10: New Sequence Table for Part Type 1

Step 5: Repeat the last two steps for the remaining three part sequence tables using the appropriate
processing times from Table 5.5.

Step 6: In order to have a single source produce multiple types of entities, a new data table has to be
created to specify the part type entities and the part mix percentages. Add a new data table called
TableParts with four columns with the specified types as seen in Table 5.6.

62 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Table 5.6: Column Definitions
Column Column Name Column Type
1 PartType Entity
2 ProcessingTimes Expression
3 PartMix Integer (or Real)
4 NumberToArrive Integer (or Real)
Step 7: Next specify each entity as a new row with the appropriate processing times, part mix percentages
and sequence table as shown in Table 5.7.

Table 5.7: Content of TableParts

Specifying SequencePart1.ProcessingTimes states that the processing time will come from the
column associated with ProcessingTimes in the table SequencePart1 for the current row.

Step 8: In order to use the dependent processing times, change each of the processing time expressions in
the three stations to be TableParts.ProcessingTime. This logic specifies that the processing time will
come from the ProcessingTimes column of the data table TableParts. As the parts move through their
individual sequence table (i.e., row to row), it will retrieve the processing time for the current row. The
property is now a reference property type which
can only be changed back to a regular property by right clicking on the label (Processing Time) and
resetting it.

Step 9: To generate multiple part types from the same source, the Entity Type needs to be changed from a
specific entity type. In the SOURCE, specify the Entity Type to come from the TableParts.PartType. This
will allow the Entity Type to come from one of the four entities specified in the table.

Step 10: Set the interarrival time of entities to be an Exponential with a mean of ten minutes (i.e., same
as a Poisson arrival rate of six per hour). Set the Entities Per Arrival to TableParts.NumberToArrive.

Step 11: To determine which specific part type is created, the part mix percentages need to be utilized. The
SOURCE has a property category called “Table Reference Assignment” which can assign a table and a row
of that table to the entity.51 There are two possible assignments: Before Creating Entities and On Created
Entity. We want to do the assignment before the entity is actual created in order to create the correct entity
type and the correct number to arrive. Therefore, specify the table shown in Figure 5.11.

51 In later chapters we will utilize Processes and the Set Row step to assign a table to an entity.
Data-Based Modeling: Manufacturing Cell 63

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Part Type

Number to Arrive

Parts Table

Row

Figure 5.11: Making the Table Reference Assignment

You specify which table (TableParts) and which row in the table to associate the object. In this case, the
row is determined randomly by using the PartMix column and specifying that a random row should be
generated (TableParts.PartMix.RandomRow).52

Step 12: Save and run the simulation model for 40 hours at a slow speed to be sure each of the parts follows
their sequence appropriately and answer the following questions.

Question 12: How long is each part type in the system (enter to leave)?
________________________________________________________________________________

Question 13: What is the utilization of each of the SERVERS?

________________________________________________________________________________

Question 14: What is the overall time for all part types in the system?
________________________________________________________________________________

Part 5.5: Using Relational Tables

The original model utilized just four part types, however we may need to add additional parts. This would
require adding a new entity, adding rows into the TableParts along with creating a new sequence table for
each part. Each addition now requires design time to continue to update the model as well as adding sequence
tables. Instead, can we drive these specifications from a spreadsheet that contains the number of parts and their
sequences? Or is there an easier way to implement the creating and sequencing of parts?

One of the inventive modeling constructs in SIMIO is the use relational data tables. SIMIO has implemented
limited database capabilities that can facilitate the modeling of relationships. Figure 5.12 shows a “relational
diagram”53 where the TableParts data table is related to the SequencePart Table via the ID column. ID is a

52
SIMIO will normalize the numbers so the probability sums to one. Therefore, you may specify whole numbers 25, 25, and 50 rather
than percentages 0.25, 0.25, and 0.5.
53 More formally it is called an Entity Relationship Diagram, but the word “Entity” is use differently than it is used in SIMIO.
64 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 primary key (PK) in the TableParts which uniquely identifies each row in the table (i.e., cannot be any
duplicates). The child table (i.e., SequencePart) inherits the primary key from the parent table (i.e.,
TableParts) which is denoted as a foreign key (FK). A particular part can have many rows (records) in the
child table.

SequencePart
TableParts

PK ID
FK1 ID
PartMix Sequence
ProcessingTimes

Figure 5.12: Setting up a Relation between Table Parts and the Sequence Part
Step 1: First, we will no longer need just four part types since we want to make this a more generic model.
Delete all but one of the part type ENTITIES and rename the remaining entity EntParts. From the
Additional Symbols section, add five additional symbols making sure to color each new entity as seen in
the rotated Figure 5.13.54

Figure 5.13: Adding Additional Symbols for the Part Entity

Step 2: For the SrcParts, set the Entity Type property to EntParts.55

Step 3: From the Data tab, delete all the columns of the TableParts except for the Part Mix and Number to
Arrive columns. Insert a new Integer Standard Property named ID. From the Edit Column section, you
can move the new column to the left and set the column as a PK using Set Column As Key button in the
same section as seen Figure 5.14. Set the ID for each of the part types to a unique number (i.e., 1, 2, 3, and
4).

Figure 5.14: Parent Table TableParts with the ID Primary Key

Step 4: Select the SequencePart1 sequence table and rename it SequencePart. From the Add Column
section, select the Foreign Key button. Set the properties for the column to specify TableParts.ID as the
primary table key to relate column too as seen in Figure 5.15. Change the name of the column to be ID.

54
Note the symbols start with “0” index.
55
Right click the property and reset it back from the reference property.
Data-Based Modeling: Manufacturing Cell 65

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 5.15: Specifying the Foreign Key Property.
Step 5: Set the ID for all the current rows to be one.

Step 6: Select the SequencePart2 tab and then highlight all the rows by selecting the first row using the
row selector and dragging down. Then copy all the rows and transverse back to the SequencePart table.
Select the last row and then paste in all the values. Change the IDs of this new part to two.

Step 7: Repeat for the last two part sequence tabs setting their IDs to three and four respectively as seen in
Figure 5.16.

Row Selector

Figure 5.16: The New Sequence Parts Table

Step 8: Once the table is set up, select the TableParts table again and observe that each of the rows now
has an expander ( ) associated with it. Figure 5.17 shows the related records for “Part type 0.” If an entity
is assigned a row from the TableParts table, it will automatically be assigned the related records in the
SequencePart table.

66 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Related Table
Records

Figure 5.17: Seeing the Related Table Information

Step 9: We need to modify the entity properties to utilize the new sequence table and id for changing the
picture symbol. Under the Animation section change the Current Symbol Index to be
“TableParts.ID-1” so the color will change appropriately.56 Also make sure the Initial Sequence
Property is set to “SequencePart” as seen in Figure 5.18. Note, the entity will only receive the related
records and not the entire SequencePart table.

Figure 5.18: Modifying the EntParts Property to Utilize the New Tables
Step 10: For each of the three server stations, change the Processing Time to be the reference property
SequencePart.ProcessingTimes rather than TableParts.ProcessingTimes since this column is now only in
the related table.

Step 11: Save and execute the model.

Question 15: What is the utilization of each of the SERVERS now?

________________________________________________________________________________

56
If we had a lot of parts, we can utilize a status label to display the ID as the part moves through the network making this work for
any number of parts. This concept will be explored in later chapters.
Data-Based Modeling: Manufacturing Cell 67

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 16: What is the overall time for all part types in the system?
________________________________________________________________________________

Question 17: Is there anything different in this model as compared to the one from the previous section?
________________________________________________________________________________

Part 5.6: Creating Statistics

SIMIO automatically produces statistics for various objects within the simulation. For example, there are
statistics on the MODELENTITY number in system and time in system or for any SERVER there is a scheduled
utilization, a number in the INPUTBUFFER station, the waiting time in the station, etc. However, in spite of
their comprehensive nature, that may be insufficient. For example, our latest model has only one entity type,
so the MODELENTITY statistics are not separated by the part type. What if we wanted to know, on the average,
how many Part 3s are in the system during simulation, and what is the average time these parts spend in the
system?

The two major statistical categories are: TALLY statistics and STATE statistics. The TALLY statistic is based
on observations like waiting time, time in system, cycle time, etc. The STATE statistic such as number in
queue, number in inventory, etc., is a time-weighted statistic such that the time a value persists is used to
compute the statistical value. While the STATE statistic may appear to be the more complicated, it is the
easiest to specify in SIMIO. For example, let’s add a STATE statistic to determine the average number of Part
3’s in the system first and then add a TALLY statistic for the average times in the system.

Step 1: Before creating a STATE statistic, it is necessary to create a state variable the will hold the values
for example the number of Part 3’s in the system. SIMIO will monitor the value of the state variable
through a STATE statistic and weight its value according to the time a particular value persists throughout
the simulation. Since the number of Part 3’s in the system is a model-based characteristic (not associated
with particular entities), we need to define a model-based state variable. From the Definitions→States
section, insert a new Discrete, Integer variable with the name GStaPart3NumberInSystem as seen in
Figure 5.19

Figure 5.19: Discrete, Integer, Model-based State Variable

Step 2: The variable needs to be incremented when a Part 3 enters the system and then decremented when
a Part 3 leaves the system. Let’s use the State Assignments of the SrcParts for Before Exiting employing
the Repeating Property Editor to increment the new state variable as shown in Figure 5.20.

Figure 5.20: Incrementing the State Variable

68 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 3: The Math.If() function simply tests if TableParts.ID has the value of “3” which refers to Part
3’s then adds “1” to the state variable, otherwise the expression adds “0.”

Step 4: In a similar fashion, the state variable can be decremented in the State Assignments of the
SnkPartsLeave for On Entering as shown in Figure 5.21.

Figure 5.21: Decrementing the State Variable

Step 5: Now all you need to do is to insert STATESTATISTIC from the Definitions→Elements section
named StateStatNumberPart3InSystem that watches the state variable as seen Figure 5.22.

State Variable being

monitored by this
statistic
Figure 5.22: Defining the State Statistic Element
Step 6: Save your model and run your simulation for 40 hours.

Question 18: What did you get for the average and the maximum number of Part 3’s in the system?
________________________________________________________________________________

Step 7: Now, we would like to collect time in system for the Part 3’s which is a TALLY statistic. The TALLY
statistic needs to record individual observations. Tally statistics can be collected at any BASIC or
TRANSFER node in their Tally Statistics property.57 Unlike the STATE statistics, the TALLY statistic needs to
be defined first. Using the Definitions→Elements section insert a Tally Statistic named
TallyStatTimePart3InSystem. See Figure 5.23 for specifying the Unit Type property of “Time.”

Be sure to set the

Unit type to
Time

Figure 5.23: Defining the Tally Statistic Element

Step 8: Select the Input@SnkPartsLeave node (i.e., the one attached to the SINK). Under the Tally
Statistics Section using the On Entering tally property, specify which observation (i.e.,
ModelEntity.TimeInSystem) to record using the Repeating Property Editor. Specify the
collection of the tally statistic as shown in
Figure 5.24.

57 In later chapters we will use the Tally process steps to collect statistics anywhere.
Data-Based Modeling: Manufacturing Cell 69

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Condition for
Collecting
Observations

Value being
collected
Figure 5.24: Specifying the Collection of Part 3 Time in System
Step 9: Note that a Custom Condition causes only the Part 3’s time in system to be collected.

Step 10: Now run your simulation for the 40 hours.

Question 19: What did you get for the average, maximum, minimum, and number of observations for time in
system for Part 3’s?
________________________________________________________________________________

Part 5.7: Obtaining Statistics for All Part Types

Since statistics like time in system and number in system would be useful for each part type, we will extend
the statistics collection to all part types rather than just Part 3’s.

Step 1: First, let’s create the time in system statistics. Navigate to the Definitions→States section, and
modify the definition of GStaPart3NumberInSystem. Rename the variable to
GStaPartNumberInSystem. Set the Dimension Type property of the state variable to a “Vector” with
“4” rows, as shown in Figure 5.25. This will create a vector of four state variables one for each part type.

Figure 5.25: Defining the Number in System Vector

Step 2: Next change the state assignments at SrcParts and SnkPartsLeave to increment and decrement
the number in system for each part type respectively as shown in Figure 5.26. Note the use of the row
(TableParts.ID) in the vector.58

Figure 5.26: Incrementing and Decrementing the Number in System

58
The row index of the first row for Vectors, Matrices, and Data Tables is one while symbol indexes start at zero.
70 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 3: You can now define state statistics for all the parts. Under Definitions→States, create three more
state statistics named StateStatNumberPart1InSystem, StateStatNumberPart2InSystem, and
StateStatNumberPart4InSystem. You need to specify the state variable for each of the statistics using
the previously defined vector state variable. This specification is the same for each statistic, except for the
row. The specification for Part 2’s is shown in Figure 5.27. Modify all the four state statistics.

Figure 5.27: Specification for StateStatNumberPart2InSystem

Step 4: Now define the TALLY statistics for the other three part typess (Parts 1, 2, and 4), just as was done
for TallyStateTimePart3InSystem (see Figure 5.23 making sure the Unit Type property is equal to
“Time”.

Step 5: Navigate to the Data→Tables section and select the TableParts DATA TABLE. Insert a new Tally
Statistic Element Reference property named TallyStatTimeinSystem, which will allow us to specify a
tally statistic for each part. Therefore each entity (i.e., part type) will know which statistic to update. There
is little need to insert State statistics into the data table since we don’t reference them within the model –
only the state variable that the state statistics monitor. Fill in the tables as shown in Figure 5.28.

Figure 5.28: Data Table Including Tally Statistics

Step 6: Once the Table row reference has been assigned, each MODELENTITY (i.e., part) will have an
associated TALLY statistic which will eliminate using complicated deciding logic to determine which
TALLY statistic to update. Select the Input@SnkPartLeave BASIC node in front of SnkPartLeave and
under the Tally Statistics entry click on the On Exited.59 Use the Tally Statistic Name will be based on the
row that has been assigned to the entity (i.e., TableParts.TallyStatTimeinSystem). The
observation Value will be the model entity time in system value as seen in Figure 5.29 to utilize the
entity’s tally statistic.

Figure 5.29: Time in System Tally Statistics

59
The On Entering could be used as well.
Data-Based Modeling: Manufacturing Cell 71

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 7: Now re-run the simulation and examine the results.

Question 20: Do the tally statistics on time in system for each part type seem correct?
________________________________________________________________________________

Question 21: Do the state statistics for number in system for each part type seem correct?
________________________________________________________________________________

Part 5.8: Automating the Creation of Elements for Statistics Collection

Adding intelligence to entities made it easy to model different routings as well as to calculate time in system
and number in system statistics for each part type without having to create multiple sink and source objects
corresponding to the individual parts. But we would still needed to define the associated statistic objects and
state variables. Such an approach becomes very cumbersome if we want to add many new part types.

An advantage of using related tables is that they may be linked (bound) directly to a spreadsheet containing
many part types and sequences. Nevertheless, we may want statistics for time in system (i.e., cycle times) and
number in system for each of the different part types. However, this situation will require a TALLY and STATE
statistic for each part type. Having to define each statistic could negate the advantages of using related tables
as individual tally and state statistic elements need to be created and then the correct statistic must be collected
for the particular part. However, SIMIO allows for the automatic creation of state variables and statistic
elements.

Step 1: If you are using the model from the previous section, you should delete all the TALLY and STATE
STATISTIC elements, so these can be created automatically.

Step 2: Since the number of part types might change, the state variable GStaPartNumberInSystem will
need to be modified to match the number of parts in the table each time. In our case, we will need to set it
equal to maximum possible number of part types in the system (e.g., 10).60 The size of the state variable
can be tied to the number of rows and columns of a data tables. From the Definitions→States section
change the Rows property from “4” to “10”.

Figure 5.30: Specifying the Size of the State Variable to Match the Table
Step 3: Navigate to the Data→Tables section and select the TableParts DATA TABLE. Notice Statistic
Element Reference property named TallyStatTimeinSystem. Remove the tally statistics from that
column, so it looks like Figure 5.31. Note that the tally statistics references are missing.

Figure 5.31: Adding a Tally Stat Reference Column

60
“Matrix from Table” is another Dimension Type for state variables which will create a two dimensional matrix based on the number
of rows and numeric columns of a table. Dynamically. However, it initializes the variable based on the table values.
72 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 4: Once one specifies entries (i.e., names), the elements will be created. The element’s initial
properties can be manually specified (e.g., Category, Data Item, Unit Type, etc.) from the
Definitions→Elements section. However, since the number of part types can be dynamic we do not want
to have to specify these properties each time, which would defeat the benefit of automatic specification of
the statistical elements. Therefore, the elements that are created can also have their properties initialized
by specifying another column in the table as their values. For our problem, the category and data item
classification of the TALLY statistics can be specified as we want it to be the same as other time in system
statistics (i.e., “FlowTime” and “TimeInSystem”). Therefore, insert two Standard String property columns
named Category and DataItem as seen Figure 5.32.

Figure 5.32: Adding Properties to the Statistical Elements

Step 5: For the Unit Type property of the TALLY statistic, insert a Standard Enumeration property column
named UnitType into TableParts, as was done in Figure 5.32. An enumeration is a property whose value
is specified from a list of potential values. In this case unit types can be “Unspecified”, “Time”, “Length”,
etc. For the Enumeration column property, specify the Enum Type to be “UnitType” to pull its values from
unit type enumeration list and the Default Value should be “Time” as seen in Figure 5.33.

Figure 5.33: Setting up the Enumeration Property Column

Step 6: Next, insert a State Statistic Element Reference property named StateStatNuminSystem as seen in
Figure 5.32.

Step 7: Rather than creating each TALLY statistic to specify as entry values, SIMIO has the ability to
automatically create elements (TALLY and STATE statistics, MATERIALS, etc.) specified from a table
column. Therefore, each part row can have a specified TALLY statistic (i.e., name) that will automatically
be created and used to keep track of its on time in system statistics. By default the Element Reference
property column will be of a “Reference” type meaning the element would need to be already created in
order to be specified as an entry in the table.

· Therefore, change the Reference Type property to “Create” which is explained in Table 5.8 and seen in
Figure 5.34.
· Set the Auto-set Table Row Reference property to “True” as well.

Data-Based Modeling: Manufacturing Cell 73

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Table 5.8: Specifying Properties to Automatically Create the Element
Property Description and Value
“Reference” - It allows you to simply specify an element that is already been defined
under the “Definitions” tab.
“Create” - Entries in this column will now automatically create a new element of the
Reference Type
specified type with the name of the entry. Note these elements will appear in the
Definitions→Elements but they cannot be deleted from here. Any changes in names
in either location (i.e., TABLE or Definitions→Elements) will be reflected in both.
Initial Property This can be used to initialize the element based on values in other columns of the
Values table.
Auto-set Table Row If “True”, then the element that is pointed to by each row will automatically be given
Reference a table reference set to that row. This has to be case if your creating and initializing
the element based on other columns

Does not
appear initially

Figure 5.34: Specifying Element Reference to Automatically Create the Element

Step 8: After specifying the “Create”, the statistical elements be created but they will not have their
properties referenced. In order to use the entries specified in these three new columns as the initial
properties, select the TallyStatTimeinSystem column and then click the Repeating Property button of the
Initial Property Values property. Next, add the following three properties and values as specified in Table
5.9.61 To specify the initial properties use the repeating property window as shown in Figure 5.35.

Table 5.9: Specifying the Initial Properties

Property Name Value
Unit Type TableParts.UnitType
Category TableParts.Category
DataItem TableParts.DataItem

Expression Editor
doesn’t work on Value

Figure 5.35: Specifying the Initialization Properties of the TallyStat Reference Column

61
You will need to type in the values directly as this is not an expression editor.
74 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 9: Select the StateStatNumInSystem and specify that this column will be created as well as seen in
Figure 5.36.

Figure 5.36: Having the State Statistics to be Automatically Created

Step 10: Next we need to initialize the new state statistics in a similar fashion as the Tally statistic column.
However, this time we will not setup additional columns to specify the parameters. Set the “Category” and
“DataItem” properties to string values and the “StateVariableName” property will use the
TableParts.ID to specify which state variable the statistic should monitor as seen in Table 5.10 and
Figure 5.37.
Table 5.10: Specifying the Initial Properties
Property Name Value
StateVariableName GStaPartNumberInSystem[TableParts.Id]
Category Content
DataItem NumberInSystem

Figure 5.37: Initial Parameters of the State Statistic Column

Step 11: Add in the specifications for the properties of the four part types as shown in Figure 5.38.

Figure 5.38: Specifying the Properties of the Statistical Elements

Step 12: You may now want to reset the Reference Type property to back to “Reference”. Then when you
set the Reference Type property to back to “Create”, the statistical elements will be automatically be
created as shown in Definition→Tally Statistical Elements (Auto-Created) section. Note how each
statistic takes its UnitType, Category, and DataItem from the DATA TABLE.

Data-Based Modeling: Manufacturing Cell 75

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 13: Now for every new row that is added, a new TALLY and STATE statistic will be created that is
initialized from the entry values of the Category, DataItem and UnitType columns. For any rows that
existed before the two columns were created, the values will not be initialized.62 We can cut and paste
values back and forth from SIMIO tables and Microsoft Excel™ spreadsheets. Select the entire table by
highlighting the row selectors and then cut the rows (Ctrl-x). Next the paste them into a Microsoft Excel™
spreadsheet so you don’t have to retype all of the information. Modify the names of the tally and state
statistics, categories, data items and the unit types as seen in Figure 5.39.

Figure 5.39: Using Excel to Modify the Values

Step 14: Copy the values in Excel and the select the empty row in the SIMIO data table by clicking on the
row selector and then paste the values into the table which should now look like the one in Figure 5.31.

Row Selector

Step 15: Save and run the model for 40 hours.

Question 22: What is the average time in the system and number in system for each of the parts?
_____________________________________________________________________________

Question 23: Does the time in system for Part3 in TallyStatistic3 agree with the previous value computed?
_____________________________________________________________________________

Step 16: Now add a fifth part type by creating a sequence and update only the data table (don’t create any
elements) by copying the fourth part information changing the tally and state statistic column values as
well as copy the fourth part sequences for the fifth part.

Question 24: Did SIMIO create the tally statistical element automatically?
_____________________________________________________________________________

Question 25: When you execute the model, did the flow time and time in system statistics for the fifth part
type appear with the proper unit type, category, and data item?
_____________________________________________________________________________

Part 5.9: Commentary

· The Start Date property of the Work Schedule intuitively would be when the schedule will start.
So if one specified a future date then nothing would happen until that date. That is not the case.
The Start Date property represents the particular day the first day of the pattern will starts. If this
date is in the future compared the start of the simulation, it will repeat backwards to the current
date based on the pattern. For example, if we have a three day pattern named Day1, Day2, and
Day3 and the work schedule start date is set to 09/17/15, schedule will follow the Day1 pattern for
this date. If the simulation starts on 09/13/15, then the pattern for 09/13 is Day3, 09/14 is Day1,

62
We can either modify the properties our self or we can delete the rows and recreate them. You can select the AutoCreate dropdown
in each column value.
76 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 09/15 is Day2, 09/16 Day3 which makes Day1 be on 09/17 and Day2 on 09/18, etc. In order to
have a work schedule not start until a particular day, exceptions to the work day have to be
employed.
· Another way to look at SIMIO properties and states is that property values are established for each
object in the Facility window at the beginning of the simulation execution and cannot be changed
during “run-time”, whereas the value of states are also established at initialization but the can be
changed during “run-time”. Objects that are created during run-time, such as ENTITY, have their
properties established during an instance of run-time which is during the initialization of that
instance, but they otherwise cannot be changed during the simulation execution.
· The use of relational tables offers a tremendous advantage over other simulation languages in
being able to assign one table and then automatically be assigned all records of related tables. We
will explore this feature more in later chapters. In later chapters, we will demonstrate how you can
bind data tables to Excel spreadsheets which could facilitate the automatic creation of new parts
and sequences.

Data-Based Modeling: Manufacturing Cell 77

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Chapter 6
Assembly and Packaging: Memory Chip Boards
A common simulation environment in production is assembly and packaging. What makes these operations
challenging is that entities are being combined together. The way they are combined can be complicated.
Sometimes the assembly is temporary and the assembly (called a batch) must be separated later. These
considerations cause us to have interest in the COMBINER and SEPARATOR object.

Part 6.1: Memory Board Assembly and Packing

Memory boards have memory chips inserted into them at an assembly operation, whose capacity is four. Each
memory board requires four chips. Memory boards as well as individual memory chips arrive randomly, so
the assembler waits for four chips of the same type and one board to be available before doing the assembly.
After the board is assembled it is sent to one of three packing stations, each of which has one worker. Boards
are sent to the packing station with the smallest queue. Figure 6.1 depicts the memory board assembly and
packing process.

Packer 1
Memory Boards Arrive Triangular(3,5,8) minutes
Exponential(2) minutes

Board Assembly
(Board + 4 chips of same type) Packer 2
Uniform(2,4) minutes Triangular(3,5,8) minutes

Choose
Smallest Line

Memory Chip Arrives

Exponential(0.5) minutes Exit
Packer 3
Triangular(3,5,8) minutes

Two Types of Chips

50% Type 1, 50% Type 2

Figure 6.1: Memory Board Assembly and Packing

The basic numerical characteristics of the assembly problem are found in Table 6.1.

78 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Table 6.1: Numerical Distributions of the Memory Board Assembly
From To Travel time (minutes)
Memory board arrivals Board Assembly 4 minutes
Memory chip arrivals Board Assembly 5 minutes
Travel
Board Assembly Packer Station 1 Pert(10,12,15) minutes
Times
Board Assembly Packer Station 2 Pert(5,7,10) minutes
Board Assembly Packer Station 3 Pert(4,5,7) minutes
Any Packer Station Exit 3.5 minutes

Interarrival time for Memory Boards Exponential(2) minutes

Arrival
Interarrival time for Memory chips Exponential(.5) minutes
Information
50% of memory chips are of type 1 and 50% are of type 2 Discrete
Processing Board Assembly processing time Uniform(2,4) minutes
Times Packing time for all Packers Triangular(3,5,8) minutes

Step 1: Create a new model that has two SOURCES named SRCMEMORYBOARD and SRCMEMORYCHIP (i.e.,
one for boards and one for memory chips). Add two MODELENTITIES into the model, one for boards
named EntMemoryBoard and one for memory chips named EntMemoryChip.63 Make sure the Entity
Type property of the two sources will create the appropriate entity type.

Step 2: Give the entity type associated with boards a different symbol as seen in Figure 6.2. You can use
one of the stock SIMIO symbols or download a symbol from Trimble 3-D warehouse like Chapter 2. You
can also make your own symbols by clicking the “Create New Symbol” menu item from the New Symbol
dropdown in the “Create” section of the “Project Home” tab. In the properties window, name the symbol
SymMemoryBoard. When creating a new symbol, you should give the symbol “height” by specifying a
value (in meters). In the case shown in Figure 6.2, a rectangle was created with a line width of 0.1 meters
and a height of 0.25 meters64. The rectangle was given a fill color of grey with a line color of green. One
should not be worried too much about the exact specifications since the symbol can be adjusted manually.

Figure 6.2: Changing the Name of the Symbol and Specifying a Height and Line Width

Step 3: Select the “Model” in the [Navigation Panel] to return back to simulation model. Select the
EntMemoryBoard model entity and then click on the SymMemoryBoard symbol in the
SYMBOLS→PROJECT SYMBOLS section to utilize our new symbol.

Step 4: For the memory chips, add an additional symbol65 since we have two different types of memory
chips. You should color the second symbol a different color (e.g., red) to distinguish them in the system.66

Step 5: Insert a COMBINER named CmbBoardAssembly into the model for the assembly operation as
shown in Figure 6.3 which will batch/combine a parent entity object with a member entity object. Set the

63 To add entities, drag MODELENTITY from the [Project Library] panel.

64
Note memory boards are not ¼ meter thick but were made this thick to show a depth in 3-D view.
65 Select the EntMemoryChip MODELENTITY and the click Symbols→Additional Symbols→Add Additional Symbol button.
66
Select the 3-D view and color the sides of the new symbol as well.
Assembly and Packaging: Memory Board Chips 79

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 initial capacity of the CmbBoardAssembly to four and the processing time to Uniform(2,4) minutes. The
combiner has two input nodes (i.e., a PARENTINPUT and a MEMBERINPUT). The parent object is the board
while the member object is the memory chip. Connect the two SOURCES via TIME PATHS to the appropriate
nodes using the time specified in Table 6.1.

ParentInput
Node

MemberInpu
t Node

Figure 6.3: Initial Memory Board Model

Step 6: Objects have characteristics that describe and /or define their behavior or attributes. Properties and
state variables are two types of characteristics that are very similar and allow the user to specify certain
characteristics of the object (e.g., processing time for a SERVER object). Properties are characteristics that
are set (initialized) when the object is created. For most objects, the initialization occurs at the beginning
of the simulation and, therefore, cannot change during the run of the simulation. State variables are
dynamic characteristics which allow their values to change during the simulation run and should be used
when their values need to be altered during the simulation. There are two object definitions that we can
modify (i.e., MODEL and the MODELENTITY) as seen in the [Navigation] panel. Both the MODEL and
MODELENTITY can have properties and state variables. However the MODEL has only one set of properties
and state variables that can be accessed (i.e., global scope) while each individual entity that is created will
contain its own properties and state variables independent of the other entity instances.67

Step 7: To make the state variables clear when they are used in various places in a model, we prefix a
MODELENTITY state variable with ESta… and a MODEL global state variable with GSta….

Step 8: Recall a board requires four identical memory chips. Therefore, the COMBINER should batch four
members of the same type. Let’s add an entity-based state variable to distinguish between the two types of
memory. Select the MODELENTITY in the [Navigation] panel and then choose the “States” section from the
“Definitions” tab. Since there are two distinct chip types, insert a new “Integer Discrete State” variable
named EStaChipType as seen in Figure 6.4.

67Both the MODEL and MODELENTITY can have a state variable named Color. The only difference is how they are accessed. Specifying
Color will access the variable of the MODEL while ModelEntity.Color will access the entity’s variable.
80 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 6.4: Specifying the Discrete Integer State Variable

Step 9: Returning to the “Facility” tab of the Model, click on the SrcMemoryChip source. To assign the
new EStaChipType variable, a Before Exiting State Assignment will be specified which makes the
assignment right before the entity leaves the source. Click on the box to the right of Before Exiting State
Assignments which brings up the “Repeating Property Editor” and add the following properties.

· Since it is equally likely the chip will be type one or type two, a random discrete distribution will be
used to assign the EStaChipType value a one or a two.

State Variable Name: ModelEntity.EStaChipType

New Value: Random.Discrete(1,0.5, 2, 1.0)68

Figure 6.5: Setting the Entity’s ChipType in a State Assignment

Step 10: Since the chip entities can be distinguished, batch four members (i.e., chips) at the COMBINER by
specifying Match Members as the Matching Rule with ModelEntity.EStaChipType as the
Member Match Expression property as seen in Figure 6.6.69

68 The Discrete distribution must use the appropriate cumulative probabilities.

69
Combiners can match any entity, match members or match certain members and parents based on a criterion.
Assembly and Packaging: Memory Board Chips 81

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 6.6: Specifying the Matching Logic for a Combiner
Step 11: Insert three SERVERS, (i.e., SrvPacker1, SrvPacker2, and SrvPacker3) one for each packer with
the appropriate processing times (see Table 6.1) and one SINK named SnkExit (see Figure 6.7). Utilize
TIMEPATHS to connect the assembly to the packers and the packers to the exit. Specify the appropriate
travel times specified in Table 6.1 for the time paths between the assembly and the packers, but do not
specify the times between the packers and the exit quite yet.

Use this
TRANSFERNODE

TIMEPATHS

Figure 6.7: Packing Portion of the Model

Step 12: Recall, the board assemblies need to choose the packer with the smallest number assigned to it. At
the output TRANSFERNODE associated with the COMBINER, the branching decision needs to be specified
using the Entity Destination Type property specification in the TRANSFERNODE.

Step 13: Once the board finishes assembly in the COMBINER it can select one of the three destinations:
Input@SrvPacker1, Input@SrvPacker2, or Input@SrvPacker3 depending on the number of already
boards assigned/designated to each packer. Create a new Node LIST (see Definitions→Lists tab) named
LstPackers, beginning with Input@SrvPacker1 as seen in Figure 6.8. Lists are necessary whenever you
need to choose from a set of objects (e.g., nodes, resources, etc.).

82 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 6.8: Node List for Selecting the Best Packer
Step 14: Now at the TRANSFERNODE, specify that entities should transfer to the destination according to the
following properties as seen in Table 6.2. Once you specify the Selection Goal property, you need to
expand it to access the Selection Expression. We need an expression that can be evaluated for the
“shortest queue”. However, do we just mean the number in the respective queues? Probably not. How
about including the number of entities being processed in the SERVERS as well as the number of entities
traveling to the SERVER.

Table 6.2: Properties for Selecting the Shortest Line

Property Value
Entity Destination Type: Select from List
Node List Name: LstPackers
Selection Goal: Smallest Value
Selection Expression: Candidate.Node.NumberTravelers.RoutingIn +
Candidate.Server.InputBuffer.Contents.NumberWaiting +
Candidate.Server.Processing.Contents.NumberWaiting
Selection Condition: Left Blank70

The CANDIDATE designation is a “wildcard” reference that it takes on the identity of the object it is
referencing. Here it refers to the node to which the entity is being routed and the particular INPUTBUFFER
and PROCESSING queue of the SERVER at the end of the path. In this example, the word CANDIDATE is
replaced by each node in the list and the expression is evaluated. In a later chapter, ASSOCIATEDSTATIONS
functions will be used to simplify this expression. Leave the Selection Condition blank and leave the
Blocked Destination Rule as the default value.

Question 1: How does the Selection Expression cause the alternative destinations for the entity to change,
relative to each destination?
________________________________________________________________________________

Step 15: Since the travel time to exit is the same from all packers and we don’t expect to change it during
the simulation, let’s specify the travel from the packers to the exit using a “Model Property” via the
“Definitions” tab. Insert an Expression property from the “Standard Property” drop-down named
TimeToExit and give it a Default Value of 3.5 with Unit Type of Time with Default Units as Minutes.
Since this is an “Expression” property then one could specify an expression for the time to exit (for
example, Random.Uniform(2,10)) while a numeric property only allows constant (i.e., 3.5).

Step 16: Now you can specify as a referenced property for the Travel Time on each of the
time paths to the exit71. Now if you want to change the travel time on these three paths at the same time,

70
It is very easy to specify the Selection Goal property as the Selection condition. The selection condition property is an expression
that has to evaluate to “True” before the item can be selected from the list (e.g., SERVER has not failed) as a possible choice.
71
Select all three time paths and then right click on the Travel Time property and select the new reference property TimeToExit.
Assembly and Packaging: Memory Board Chips 83

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 only the new property has to be changed or you want to experiment on the impact of this time. You can
access the model properties by right clicking on the MODEL in the [Navigation] and then expanding the
“Controls” category to access our new property as seen in Figure 6.9.72

Step 17: The MODEL properties become controls to the simulation. These can be changed at the beginning of
any run. Furthermore they become changeable scenario properties in experiments as we saw earlier in
Figure 6.9.

Figure 6.9: Changing the Properties of the Model

Step 18: Save and run the model for 10 hours answering the following questions.

Question 2: How long do the entities stay in the system (enter to leave)?
________________________________________________________________________________

Question 3: What is the utilization of the CmbBoardAssembly?

________________________________________________________________________________

Question 4: What is the utilization of each of the packers?

________________________________________________________________________________

Part 6.2: Making the Animation Reveal More Information

Let’s first fix the memory board animation so the “chips” look attached to the board.

Step 1: Select the EntMemoryBoard MODELENTITY and insert the BATCHMEMBERS queue from the
“Attached Animation” in the “Draw Queue” drop down. Position it on top of the EntMemoryBoard as
seen in Figure 6.10.

72
Note that if a default value is not given, the simulation model may not run until you have set the property because it will be Null.
84 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 This is the
BATCHMEMBERS
ANIMATED QUEUE

Figure 6.10: Added Batch Queue

To put an object on top of another you need to select the object and hold down the Shift key to stack the
queue on top of the memory board symbol. It’s probably best to do this while switching back and forth
from 2-D to 3-D.73

Step 2: Now let’s use the color of the memory chips symbol to designate the two different chip types. To
do this, click on the MEMORY model entity and click on “Add Additional Symbol” in the “Additional
Symbols” section of the” Symbols” tab. You can color one symbol red and the other green by simply
selecting one of the “Active Symbol” and changing its color. Please note that these two symbols are
numbered zero and one (not one and two).

Step 3: Next, we need to “assign” the colors to the different chip types. Select the SrcMemoryArrivals
SOURCE. Click on the box to the right of Before Exiting State Assignments, since we want to assign the
color before the entity leaves the source. This brings up the “Repeating Property Editor”. Add an
additional property with the following characteristics.

State Variable Name: ModelEntity.Picture

New Value: ModelEntity.EStaChipType-174

Figure 6.11: Changing the Chip Entity to have a Different Color

A “state variable”, unlike a property, can be changed throughout the simulation and here the picture is a
SIMIO defined characteristic of the entity (in some simulation languages called an attribute of the entity).

Question 5: Why do we subtract one from the EStaChipType property of the entity?
________________________________________________________________________________

Step 4: Run the model to be sure the animation behaves as expected.

73
Remember that hitting the “h” key in the model window brings up instructions for moving around in 2D and 3D and also in
manipulating objects. Hitting the “h” key again removes the instructions.
74
Recall the picture symbols start at zero which is why we need to subtract one.
Assembly and Packaging: Memory Board Chips 85

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 6: Do the entities change color corresponding to their chip type?
________________________________________________________________________________

Step 5: Suppose now we want the symbol leaving the packing station to represent a “package”. There are
several ways one might do this. One easy way is to use the Before Exiting State Assignments at the
packing stations. Now at one of the Packing stations, bring up the Repeating Property Editor and this
time add:

State Variable Name: ModelEntity.Size.Height

New Value: 1

Step 6: You may need to experiment with the value. The idea is to change the memory board’s “height” so
that it encases the “chips” so they cannot be seen (i.e., hides the batch members inside the board).

Step 7: Run the model and adjust the “height” so it shows a “package” leaving the packer. You will need
to add this change in “height” at each packer station, when the entity exits.

Step 8: Look at the model in 3-D and notice the changes.

Step 9: Save the model as Chapter 6.2.spfx and run the model for 10 hours.

Question 7: How long are the entities in the system (enter to leave)?
________________________________________________________________________________

Question 8: What is the utilization of the CmbBoardAssembly?

________________________________________________________________________________

Question 9: What is the utilization of each of the packers?

________________________________________________________________________________

Question 10: What is the average number in the MemberInputBuffer?

________________________________________________________________________________

Question 11: What is the average number in the ParentInputBuffer?

________________________________________________________________________________

Question 12: What is the average number in processing queue in the CmbBoardAssembly queue?
________________________________________________________________________________

Part 6.3: Embellishment: Other Resource Needs

Sometimes capacity is not just limited at objects like SERVERS and COMBINERS. The limitation may not even
be fixed at a particular location. For example, suppose the actual board assembly requires a particular fixture
in order to assemble the chips to the board. Furthermore that fixture is used to transfer the assembly to
packing. For each assembly, the fixture must be first obtained from the packing operation, which takes about
three minutes which we will relax the assumption. Let’s assume there are ten identical fixtures available for
assembly-packing and once they are released at the pacing they can be used immediately at the assembly.

86 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 1: First, drag and drop a RESOURCE object75 onto the modeling canvas, maybe just below the
COMBINER as shown in Figure 6.12. Name the resource ResFixture and specify an Initial Capacity of 10
(we won’t be changing its capacity in this case).76

This is the Fixture

RESOURCE

Figure 6.12: Adding the Fixture Resource

Next, logic needs to be specified on how this RESOURCE is to be utilized. The fixture will be “seized” just
before assembly operation begins and then released just after packing operation finishes.

Step 2: For the CmbBoardAssembly COMBINER, select the Secondary Resources and then the Other
Resource Seizes and then the Before Processing as seen in Figure 6.13, Click the ellipses box to bring up
the Repeating Property Editor for seizing resources and request one unit of the ResFixture resource
capacity to be seized. We won’t worry about which fixture is seized, since they are identical.

Request a Fixture
Before Assembly
Starts

Figure 6.13: Seizing One Unit of Resource Capacity

Step 3: When the package has been finished, the fixture needs to be released for the next assembly. This
process has to be done for each of the three packers. For each packer, select the Secondary Resources and
then the Other Resource Releases and then the After Processing. Click the ellipses box to bring up the
Repeating Property Editor for releasing resources. As shown in Figure 6.14, one unit of the ResFixture
resource capacity will be released.

75
Later chapters will go in more detail on the RESOURCE object.
76Note, a RESOURCE is fixed object and cannot travel throughout the network. Other types of resources (i.e., WORKER, VEHICLE, and
ENTITY) can be dynamic and move through out a network.
Assembly and Packaging: Memory Board Chips 87

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Release the Fixture
After Packaging
Finishes

Figure 6.14: Releasing the Fixture

Step 4: In order to “see” how the available capacity of the resource changes during the simulation, add a
Status Label from the Animation section in the ribbon. Make sure no objects are selected.77 When the
status label is selected, the cursor can be used to draw a box onto the modeling canvas as shown in Figure
6.15. Note that the expression gives the remaining capacity of ResFixture.

ResFixture
Status Remaining
Label Capacity

Figure 6.15: Adding a Status Label for Remaining Capacity

Step 5: Run the model and record the time in system for boards and chips.

Step 6: Make the capacity of the ResFixture equal to “20” and re-run the model.

Question 13: What is the difference in the time in system for boards and chips for 10 versus 20 fixtures?
________________________________________________________________________________

Question 14: Does it make sense to purchase ten additional fixtures?

________________________________________________________________________________

Part 6.4: Changing Processing Time as a Function of the Size of the Queue

Suppose you find out that SrvPacker1 changes its processing time depending on the number currently waiting
(i.e., size of the queue). Specifically, the packer is only 80% as efficient when there is zero in the input buffer
queue, 90% when there is one, and 100% for all others.

Step 1: You need an “efficiency” table now to determine the processing time. This table can be modeled
as a “Discrete Lookup” table in SIMIO (i.e., f(x)). From the “Data” tab, add a “Lookup Table” whose
values are 0.8 for 0, 0.9 for 1, 1 for 2, and 1 for 100 as shown in Figure 6.16. Name the table
LookupEfficiency.

77 If we had selected the ResFixture first before inserting the status label, the label would be attached to the resource and therefore the

expression would be just Capacity.Remaining. If we moved the Resource in the model the status label would move as
well.
88 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 6.16: Lookup Table for Efficiency Calculation
Remember the table gets interpolated for values in between the discrete points. To access the efficiency
based on the number waiting in the queue at the SrvPacker1, we will use the following expression:
LookupEfficiency[SrvPacker1.InputBuffer.Contents.NumberWaiting]. Notice, to
access rows into (i.e., indexing) into tables and arrays, SIMIO uses the [] notation rather than ().

Step 2: We are planning on using the expression more than once in our model. Since the expression is very
long and complicated, we will utilize SIMIO’s function expression creator to alleviate mistakes.78 From
the “Definitions” tab under the Functions (f(x)) section, insert a new FUNCTION named FuncEfficiency.
Set the Expression property to the appropriate value as seen in Figure 6.17.

Figure 6.17: Specifying a SIMIO Function Expression

Step 3: Now in the Processing Time property for the SrvPacker1, change the expression such that the
Triangular distribution is divided by the efficiency function to get the actual processing time. The new
expression is defined as the following.

Random.Triangular(3,5,8)/FuncEfficiency

Question 15: Why do we divide the original processing time by the efficiency?
________________________________________________________________________________

Step 4: We want to be sure our “efficiency” function is working correctly. To do that, we will use a
SIMIO “Status Plot” which you will find under the “Animation” tab in the ribbon. Use the following
specifications for the plot which are in the “Labels” and “Time Range” section of the “Appearance” tab.

78
If we were to change the expression, it will only need to be modified in one place rather than in every location we used the direct
expression.
Assembly and Packaging: Memory Board Chips 89

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Title: Efficiency at Packer 1
X Axis: Time
Y Axis: Number in Packer 1 Queue
Text Scale: 1
Time 10 Minutes
Range:
Figure 6.18: Setting the Properties of the Status Plot
Also, add the Expression property of the Status Plot that will be drawn which is FuncEfficiency where
the plot is given in Figure 6.19.

Figure 6.19: Efficiency Plot

Question 16: Does it appear that the efficiency function is working?
________________________________________________________________________________

Question 17: What happens to the chips when they are assembled at the combiner?
________________________________________________________________________________

Question 18: Can you visually distinguish between the two chip types?
________________________________________________________________________________

Part 6.5: Creating Statistics

Here the SIMIO “automatic” statistics collection process may not provide the statistics that interest you. For
example, we might want the time the boards and chips get from entry to exit or we might want to know what
the average efficiency is for SrvPacker1. Time in system or time in a “sub-system” is an example of
“observation-based” statistics. Recall this type of statistic is referred to in SIMIO as a “Tally Statistic”. In
the case of the efficiency, its value is changing (discretely) with respect to time and it is an example of what is
generally referred to as a “time-persistent” or “time-weighted” statistic. In SIMIO this type of statistic is a
“State Statistic” and since efficiency changes instantaneously (discretely) it is referred to as a “Discrete State
Statistic”.79

Step 1: Consider first the “state” statistic “efficiency”. To collect a statistic on this characteristic, we need
to define a model-based (global) state variable for efficiency. From the “Definitions” tab, select “States”
and in the “Discrete” section click on “Real” to define a new state variable named
GStaPacker1Efficiency.

79
Don’t confuse a “state variable” with a “state statistic”.
90 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 19: Notice there are seven general discrete types of “state” variables, but only the
“Discrete→Real” one is relevant here. What is there about the efficiency change that is “discrete” and
“real”?
________________________________________________________________________________

Question 20: What are some of the other “value types” within the “Discrete” variable type?
________________________________________________________________________________

Step 2: To ensure that StaPacker1Efficiency is correctly monitored for its value by SIMIO, let’s modify
the “State Assignments” in the SrvPacker1 SERVER. In particular, we will make the assignment On
Entering the server and use the following assignments.

State Variable: GStaPacker1Efficiency

New Value: FuncEfficiency

Step 3: Finally (this is the last step in collecting a state statistic), we need to define a “State Statistic”.
Defining new statistics can be done from the “Definitions” tab, but in the “Elements” section. Elements
represent special objects that add modeling flexibility. Let’s call the new statistic the
StateStatPacker1Utilization and use GStaPacker1Efficiency as the State Variable Name as seen in
Figure 6.20.

Sets Results “Category”

and “Data Item”

Figure 6.20: Setting a State Statistic

Step 4: Save and run the simulation model for ten hours. Filter the results of the “Object Type” to “Model”
to see the new statistic.

Question 21: What did you get for the average StateStatPacker1Utilization?
________________________________________________________________________________

Step 5: Next let’s obtain the “time in system” statistics for the boards and chips to go from entry to exit.
There are several ways to do this. Although you might think you need to give some special instructions to
SIMIO, this is a case where we can take advantage of the way SIMIO provides statistics. Recall that we
automatically get “time in system” statistics on the entities exiting through a sink. We will exploit this
modeling approach.

Step 6: To obtain the statistics on the boards and chips, we can separate them and send them through
separate exits (i.e., SINKS) – and do all this without changing the fundamental statistics. Our approach is
shown below in Figure 6.21.

Assembly and Packaging: Memory Board Chips 91

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Connectors

Branch Members

Figure 6.21: Separating Entities

The approach will substitute a SEPARATOR (from the [Standard Library]) for the original SINK and then
branches the “parent” and “members” output to the appropriate SINK. The members get further branched to
separate SINKS. CONNECTORS are used to connect objects so there is no time taken.

Step 7: Again, using time paths connect the packers to the input of the SEPARATOR specifying the
TimetoExit property as the travel time. The branching of members uses “Link Weight.” The Selection Weight
property for the link connecting to the SnkRedLeaves SINK is ModelEntity.EStaChipType==2 while
the Selection Weight property for the connector to the SnkGreenLeaves SINK is
ModelEntity.EStaChipType==1.

Step 8: Run the model for 10 hours and answer the following questions.

Question 22: How long are boards in the system from the time they enter to when they leave?
________________________________________________________________________________

Question 23: How long are “red” chips and “green” chips in the system (enter to leave)?
________________________________________________________________________________

Question 24: What is the average utilization of Packer 1 over the 10 hours?
________________________________________________________________________________

Part 6.6: Commentary

· Good animation can add to modeling insight as well as providing a communication vehicle

92 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Chapter 7
Using SIMIO Processes
Most simulation languages hide the details of executing the processes associated with an event. Generally,
SIMIO has taken care of performing these steps and we have been somewhat ignorant of what specifically
happens unless you have looked at the trace. Also, it is usually difficult to look “under the covers” in most
simulation languages as well as have the ability to add your own features to the simulation model.
An innovative feature of SIMIO is the capability to modify an object’s behavior through the use of SIMIO
“Processes.” In essence you can determine some new things to happen based on current conditions. SIMIO
processes offer a wide range of modeling opportunities, which will be employed in this and later chapters.
Processes are very flexible and can model many varied behaviors, well beyond those that are intrinsic to the
object (such as a SERVER). Processes allow you to extend and expand the modeling capability of native
SIMIO. There are two basic considerations when using and developing processes.
1. “When” is the process invoked or executed?
2. “How” do you write a process to produce the desired behavior?
By now you realize that a simulation executes by moving from event to event. SIMIO has defined certain
events from which you can respond with a process changing the models behavior. Processes are a sequence of
simulation “steps” typically invoked by actions of objects within the simulation or by events whose response
you write. Writing a process is similar to creating a stylized programming flowchart, except that the
components of the flowchart are logical simulation “steps” such as Assign, Decide, Tally, Transfer,
Seize, Delay, Release, and so forth. The actual execution of the process is performed by a “token.” The
TOKEN is a SIMIO object that can have state variables but it is not the same as an entity. The tokens execute
the steps of a process on behalf of an object, typically the object whose behavior is being modified by the
process or where the process resides (i.e., PARENT OBJECT) or the object causing the modification (i.e.,
ASSOCIATED OBJECT). The relationship among the token, objects, and process is shown in Figure 7.1.

Part 7.1: The Add-On Process

We will initially focus on the SIMIO “Add-on Process” triggers. These are events triggered by actions of
objects but mostly by entities to change the behavior or cause an action for an object that is needed for your
particular model. In the case of an Add-on process, the PARENT OBJECT is primarily the main MODEL since the
Add-on processes are created within the main Model and are viewed and edited in the “Processes” tab of the
MODEL. Often, the ASSOCIATED OBJECT is the MODELENTITY that triggers the particular process to be
executed.

Figure 7.1 shows an example of the “Entered” Add-on process trigger. When Entity30 enters the SrvPacker1
server, the SrvPacker1_Entered process is triggered. At this point, a TOKEN associated with Entity30 is
created at the Begin endpoint and travels step to step in zero time (e.g., the TOKEN is currently at the begging
getting ready to execute the Assign step) until it reaches the End endpoint at which time it is destroyed. In
this example, the ASSOCIATED OBJECT is Entity30 and the PARENT OBJECT is the main Model. The TOKEN is
executing the steps on behalf of Entity30.80

80
You can have more than one TOKEN associated with the same MODELENTITY which will be seen in later chapters.
Using SIMIO Processes 93

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Associated
with Entity 30

Token

Entity 30 Enters SrvPacker1

Invokes SrvPacker1_Entered
Add-on Process Trigger

Figure 7.1: Example Explaining Entity Triggering an Add-on Process

Part 7.2: The Add-On Process Triggers: Illustrating “Assign” Step

Almost all SIMIO objects have “Add-On Process Triggers.” These triggers cause external processes to be
invoked as part of other processes. So as a part of the execution of the object, arbitrary “processes” of your
design can be also executed which helps change/expand the underlying object’s behavior (i.e., SERVER)
without having to look inside the object which will be explored in a later chapter.

In Chapter 6, the characteristics of the entities were changed. In one case, the entity types (i.e., the memory
chips) were colored and in the second case the memory board entity’s height was changed. In both cases, we
used “state variables” whose values were changed using the “State Assignments” properties of the SOURCE and
SERVER objects. You will see that there are only a few opportunities within an object to make state
assignments, while there are many more Add-On Process Triggers. Furthermore, the “State Assignments”
can only assign state variable values, whereas a process can incorporate more logic than just simple
assignments.

Step 1: Let’s see how to replicate the “State Assignments” using Add-On Process Triggers. Bring up the
model from the prior chapter. First, remove the “State Assignment” from the SrcMemoryChip SOURCE by
deleting them within the Repeating Property Editor.

Step 2: Make sure the SrcMemoryChip is selected and click on the “Add-On Process Triggers” menu to
expand it.

Question 1: What Add-On Process Triggers are available in the SOURCE object?
________________________________________________________________________________

Question 2: When “State Assignments” are performed on entities made in the SOURCE object?
________________________________________________________________________________

Step 3: Choose the “Exited” Add-On Process trigger. Double clicking on this trigger automatically creates
a new process named SrcMemoryChip_Exited and then selects the “Processes” tab. You will see a
“flowchart” with a “Begin” and an “End” as seen in Figure 7.2. By clicking on one of the process Steps,
you can insert it into the flowchart, which will adjust itself automatically.

Question 3: Name a few of the steps available that you might use in creating a process?
________________________________________________________________________________

Question 4: Why did you choose these?

________________________________________________________________________________

Step 4: Insert an Assign step as shown in Figure 7.2.

94 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Insert “Assign” Step

Figure 7.2: Inserting the Assign Step

Step 5: Under the properties window you will see a place to put the State Variable Name and its New
Value. You can only assign one state variable. However, you can also invoke the Repeating property Editor
by clicking beside the “Assignments (More).” Add the following two assignments using the latter approach as
seen in Figure 7.3.81 It should look very familiar as the “State Assignments” utilize an Assign step
internally.82
State Variable Name: ModelEntity.EStaChipType
New Value: Random.Discrete(1, 0.5, 2, 1.0)
State Variable Name: ModelEntity.Picture
New Value: ModelEntity.EStaChipType - 1

Figure 7.3: Repeating Property Editor for Assignment Process Step

Step 6: Recall from the previous chapter, we calculated statistics on the efficiency of packer one. The
efficiency was used to change the processing time of the packer based on the number in the queue. The
state variable was updated when the entity entered the packer including the current entity rather when the
entity entered processing where the efficiency value is used. The reason for not calculating the statistic
correctly is there are only two default places for assignments (i.e., entering or exiting the server) but
processes allow more flexibility. To fix the issue, delete the state assignment from the “On Entering” at
SrvPacker1. Create a “Processing” Add-On Process Trigger which will calculate the efficiency right
before the processing starts. Insert an Assign process step that does the following – see Figure 7.4.83

State Variable Name: GStaPacker1Efficiency

New Value: FuncEfficiency

81
When performing more than one assignment, it is recommended you perform all of them in the Assignments (More) to avoid issues
of duplication, etc. The single assignment is performed before the Assignments (More).
82
In the later chapters on sub-classing we will explore how SIMIO does the assignments when we take apart some of the basic objects.
83
Utilize the single assignment for this exercise.
Using SIMIO Processes 95

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 7.4: Recreating State Efficiency Assignments
Step 7: Save the model and run this model to make sure it replicates the behavior from the previous
chapter.

Question 5: What was the average efficiency and was it much different than the previous chapter?
________________________________________________________________________________

Part 7.3: Creating an Independent “Reusable” Process

It is useful to be able to create a “generic” process that can be used in several places within a model. In the
model from the prior chapter, the assignment to change the height of the board to make the entity look like a
package was done at each packer station. Of course, this approach could be duplicated by creating an “Exited”
add-on process trigger for each packer. However, the same exact logic of assigning the height has to go in
each of the three “Exited” add-on process triggers. Instead, let’s create a “common” process which we can
refer to in each of the packers, rather than creating the same process for each packer. This type of process is
called a “reusable” process and allows for modification in one place rather than three.

Step 1: To do this, select the “Processes” tab and click the “Create Process” button naming the new
process Packaged. Next add an Assign step to its flowchart with the following assignment.
State Variable Name: ModelEntity.Size.Height
New Value: 1 (Meters)
Step 2: Delete the Before Exiting “State Assignments” at each of the packer stations.

Step 3: For each of the packer object’s “Exited” Add-On Process Trigger, instead of creating a new one;
just select the Packaged process from the drop-down list. Again select all three packers to change them all
at once as seen in Figure 7.5.

Figure 7.5: Utilizing a Common Process in all Three Packers

96 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 6: What is the advantage of have one process and referring to it at each packing operation over
having duplicated processes at each packing station?
________________________________________________________________________________

Step 4: Run this model and make sure it replicates the previous behavior.

Part 7.4: Collecting Tally Statistics

Recall, observation-based statistics are referred to as TALLY statistics in SIMIO. These statistics are gathered
as observations as the simulation executes. Examples would be queuing times, flow times, cycle times, and
time in system. For the most part, these are statistics whose observations are intervals of time. Previously,
TALLY statistics collection was limited to BASICNODES or TRANSFERNODES or the default results. Now we want
to broaden greatly the possible collection points, namely in any PROCESS.

Step 1: Let’s consider collecting TALLY statistics on the time boards and chips spend in the system from
arrival time to the time they start to be packed at Packer1 only. To accomplish this addition, we need to
first define a new “Element.”84 (When working with TALLY statistics, we define them before we determine
their method of collection.) Under the Definitions→Elements section, click on “Tally Statistic” in the
ribbon to insert a new statistic named TallySystemTimePacker1 as seen in Figure 7.6.

Figure 7.6: Specifying a new Tally Statistic in the “Elements” Section

Step 2: Be sure to make the “Unit Type” for the statistic “Time”.

Step 3: Next, we need to define the collection process (i.e., logging each observation). This done by
inserting a Tally step into the SrvPacker1_Entered add-on process trigger as seen in Figure 7.7.
TallyStatistic Name: TallyStatSystemTimePacker1
Value Type: Expression
Value: Run.TimeNow-ModelEntity.TimeCreated85

Figure 7.7: Using a Tally Step to Track Time before Packing Station 1
Question 7: Could the Value expression simply be ModelEntity.TimeInSystem?
________________________________________________________________________________

Step 4: Save and run the model observing the results.

84 Recall, Elements are behaviors of objects that can change state over time.
85
Run.TimeNow is the current time of the simulation and TimeCreated function returns the time the entity was created.
Using SIMIO Processes 97

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 8: What did you get for the arrival time for boards and chips from the time they arrive to the time
they start to be packed at Packer1?
________________________________________________________________________________

Step 5: Suppose now we want only the time in system for assemblies as they start to be packed at
SrvPacker1 after the first four hours of the simulation? We need to modify our process so only
observations of boards are collected after the first four hours.

Step 6: Modify our “SrvPacker1_Entered” add-on process by adding a Decide step before the Tally
as shown in Figure 7.8. The Decide step is similar to an If statement in Excel where when the condition
is true the process (i.e., TOKEN) will continue via the “True” branch otherwise it will follow the “False”
branch. The Decide Type property should be “ConditionBased” with the Expression property set to
Run.TimeNow > 4.

Figure 7.8: Adding a Decide Step

Step 7: Save and run the model.

Question 9: What did you get for the arrival time for boards and chips from the time they arrive to the time
they start to be packed at SrvPacker1?
________________________________________________________________________________

Step 8: It is also possible to use process steps to replace the modeling changes we made to the original
model in order to produce time in system statistics for each chip type (i.e., red and green) and board. The
previous approach placed additional objects (i.e., SINKS) in the model, which were not be part of the real
system and might confuse management (or anyone else) when looking at the model.

· First, insert the original SnkExit (SINK).

· Delete all the additions (i.e., SEPARATOR and the three SINKS) when the original SnkExit SINK was
replaced.86
· Next, connect the three packing stations back to the SINK as before, making sure to specify the
TimeToExit property as the Travel Time.

Step 9: Next let’s define three new TALLY STATISTICS (remember these are “Elements”) named
TallyStatBoardTimeInSystem, TallyStatRedTimeInSystem, and TallyStatGreenTimeInSystem.
Make to change the Unit Type to Time. To make these easier to find in the output, under “Results
Classification”, set the “Category” to FlowTime, and the “Data Item” to TimeInSystem as seen in Figure
7.9.

86
When you delete a node, all the connections to it disappear. You can quickly connect nodes by holding the Shift and Ctrl key down,
and clicking on the origin node and dragging a connector to the destination. You will need to select the type of connection (here we
used TIMEPATHS whose TravelTime is a reference “TimeToExit”.
98 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 7.9: Specifying Time in System Tally Stats
Step 10: We will now create an Add-On Process in the SINK using the “Destroying Entity” add-on process
trigger. This process is called SnkExit_DestroyingEntity and its process is shown in Figure 7.10.87

Figure 7.10: Destroying Entity Process (Tallying Statistics)

Note, the UnBatch step comes from the “All Steps (A-Z)” category in the steps panel. The UnBatch
works much like the SEPARATOR object except it breaks up the parent and the members separately in the
batch and sends them along the correct branches of the step.88 Here two tokens are active on behalf of the
same associated object: one TOKEN executes on behalf of the PARENT and the second TOKEN executes on
behalf of the MEMBER89.

· The Decide step is conditional, based on the condition: ModelEntity.GStaChipType == 1

· The Tally steps simply record ModelEntity.TimeInSystem at each step for each of the
three appropriate TALLY statistics.
· The Destroy steps are needed to eliminate the “unbatched” members. The Parent will be destroyed
at the sink, but the members are unattached and will remain in the model90 unless removed.

Question 10: What are the advantages and disadvantages of this approach versus the original model where
the symbolic model was changed to gather the different statistics?
________________________________________________________________________________

87Double clicking on the Exited Add-on process trigger property will automatically create the process as well as take you to the
“Processes” tab.
88
Internally the SEPARATOR utilizes an UnBatch step to separate the parent and member entities and then sends them out the
appropriate output nodes.
89
In this case the second token will cause four tallies – one for each of the four chips. Furthermore, the execution of the “member”
occurs before the execution of the “parent” from an Unbatch.
90
The members are unbatched into “FreeSpace”, which designates nowhere in particular, but they take up computer memory.
Using SIMIO Processes 99

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 11: What are some of the other steps that appear in “All Steps (A-Z)” that don’t appear in the
“Common Steps”?
________________________________________________________________________________

Step 11: Save and rerun the model observing what happens as compared to the previous model.

Question 12: How long are the boards in the system (enter to leave)?
________________________________________________________________________________

Question 13: How long are the “red” chips in the system (enter to leave)?
________________________________________________________________________________

Question 14: How long are the “green” chips in the system (enter to leave)?
________________________________________________________________________________

Step 12: The results should be comparable to what you obtained in the previous chapter.

Question 15: Since using processes duplicated what we did with the original model, which approach is the
easiest for you? 91
________________________________________________________________________________

Part 7.5: Manipulating Resources

Previously, we used a fixture from the assembly through the packing. A RESOURCE object with a capacity of
ten was used to model the availability of the fixtures. We employed the SERVER’S built in “Secondary
Resources” to seize one unit of resource capacity before processing at the board assembly. We did not release
that capacity until the entity had finished processing at the packer station.

Unfortunately, if additional resources are needed at a SERVER or COMBINER, the places where capacity may be
“seized” or “released” is limited. Furthermore, all that can be done at those locations is a “seize” or “release”
(seize and release if used for only for “processing”). Recall, before the fixture can be used it has to be
retrieved from the packaging area prior to processing and then torn-down after packaging (i.e., packaging
separated from the memory board). One approach would be to add server objects for these activities and place
them into the model. These additional objects again begin to clutter the model with substantial objects that
perform only simple functions which processes can handle as well as model these situations easier.

Whenever we are modeling the use of capacity within an object, there are three steps that are often employed
(i.e., Seize, Delay, and Release). Using these steps you “seize”, “delay” for some time, and then
“release” capacity. When you seize capacity it is important to know how many units of a resource is being
seized. It is possible to seize units of multiple resources at the same time.

Now, we want to delay a Uniform(1,3) minutes after seizing the resource to setup and retrieve it and then
delay a Uniform(2,4) minutes for a tear down separation process before releasing the resource. Before we
simply choose one of the Add-on processes, let’s consider the choices related to the processing as seen in
Table 7.1.

91
People often try to avoid using processes at first, especially if they can model without them (this is part of the “different thinking”
required in SIMO). However you will find that the time invested in learning to use processes will pay off handsomely as you
encounter more complex systems. The remaining chapters will illustrate why processes are so essential.
100 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Table 7.1: Few of the SERVER’S Add-on Process Triggers

Add-on Process Trigger Description

Before Processing Occurs when an entity has been allocated SERVER capacity, but before entering
(i.e., or ending transfer into) the processing station.
Processing Occurs when an entity has been allocated SERVER capacity, ended transfer into
the SERVER’S processing station, and is now about to start the processing time.
After Processing Occurs when an entity has completed the processing time and is about to
attempt its exit from the SERVER’S processing station.
Exited Occurs when an entity has exited the SERVER object
We will (somewhat arbitrarily) choose Before Processing for the place to seize the fixture and After
Processing for when it should be released. Note that the delay times setting up and tearing down the fixture
therefore contributes to the time being processed at the objects.
Step 1: First, remove from the CmbBoardAssembly the “Secondary Resources/Other Resource
Seizes/Before Processing” the seize of one unit of capacity of ResFixture by deleting the entry in the
Repeating Property Editor.

Step 2: Next, remove from each of the Packing stations the “Secondary Resources/Other Resource
Releases/after Processing” the release of one unit of capacity of ResFixture by deleting the entry in the
Repeating Property Editor.

Step 3: For the CmbBoardAssembly COMBINER, create the “Before Processing” add-on process trigger
named CmbBoardAssembly_BeforeProcessing. This trigger is executed right before the COMBINER is
ready to start processing on the assembly operation. The Seize step will be used to specify that one unit
of capacity of ResFixture is needed at processing as shown in Figure 7.11. Once a fixture has been
seized, a Uniform(1,3) minute Delay will be the time to setup and retrieve the fixture.

Figure 7.11: Seizing One Unit of Fixture Capacity and Delaying

Step 4: When the package has been finished, the fixture needs to be removed and released for the next
assembly. This process has to be done for each of the three packers, so it will be easier to create a
reusable process which is invoked when the board exits the packing stations. From the “Processes” tab
click the “Create Process” button naming the new process RemoveFixture. The Delay step and the
Release step are shown in Figure 7.12. The delay will take a Uniform(2,4) minutes to tear-down the
fixture before the fixture is released.

Using SIMIO Processes 101

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 7.12: Tear-Down and Release the Fixture
Step 5: Now select the After Processing Add-On trigger of each of the Packers and select the
RemoveFixture from the drop down list.

Step 6: Run the model for 8 hours.

Question 16: After looking at the results, what concerns do you have?
_______________________________________________________________________________

Question 17: What changes to the system could we suggest to make it more reasonable?
_______________________________________________________________________________

Part 7.6: Tokenized Processes

The previous section showed how the exact same process could be reused in multiple locations. However, if
the process was not completely identical (i.e., the tear down time depended on which packer). A process may
be given arguments through the tokens that execute them. These execute in a fashion similar to arguments in a
programming language subroutine or subprogram. They parameterize the process or what we call a
“tokenized” process since it requires custom tokens.

The need for a tokenized process occurs whenever you have several almost identical processes. For example,
the time in system statistics for assemblies as they start to be packed at SrvPacker1 after the first four hours of
the simulation was needed. The “solution” was to create a process specialized to computing the time in
system to that packer. If we wanted that same time in system for SrvPacker2 and SrvPacker3, we would
need to create two new processes, one for each packer, along with two more statistics. To do this we would
probably copy the first packer process for the second two and edit them accordingly. Of course every time,
one copies and edits something, there is an increased chance of making an error. Also, if later a change or
additional behavior is needed, all three processes would need to be updated.

Step 1: Let’s instead consider using a Tokenized Process that passes-in the particular statistic to record the
operation. First, define new time in system (TALLY) statistical elements for Packer2 and Packer3 as seen in
Figure 7.13 and do not forget to change the Unit Type property to “Time.”

Time in System
statistics for Packers
2 and 3

Figure 7.13: Tally Statistics for Packers 2 and 3

102 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 2: Recall a TOKEN is created in order to execute the steps of a process. Therefore, parameters are
passed to the process via a custom TOKEN. From the Definitions→Token section, add a custom TOKEN
named TknPacker. This TOKEN needs an additional “Element Reference” state variables92 named
TStaTallyStat, to reference the correct statistic that will be used to store a parameter as seen in Figure
7.14.

Figure 7.14: TknPacker State Variables

Step 3: Rename the SrvPacker1_Entered add-on process to SrvPacker_Entered and specify its
properties as seen in Figure 7.15. Change the type of TOKEN that will execute the process steps to be
TknPacker. Then specify the Input Arguments property with the one row (input arguments) as defined in
Figure 7.16which specifies what the input parameter will be named and how it is linked to the customized
token.93

Figure 7.15: Properties of the SrvPacker Process

Figure 7.16: Input Arguments for the Process

92
Tokens always have the state variable ReturnValue.
93
Notice, you can also specify the process return values which might be useful if one process calls another process.
Using SIMIO Processes 103

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 4: Finally at each packer invoke the SrvPacker_Entered by specifying its own argument. Figure 7.17
shows the example for Packer 2.

Figure 7.17: Invoke the Tokenized Process

Step 5: Execute the model and note the tally time in system statistics for each of the three packers.

Part 7.7: Commentary

· The process capability in SIMIO is one of its standout features, allowing for extension and
customization in powerful and unique ways. To use it requires a potential change in thinking towards
simulation modeling.
· A lot of different modeling can be done with resources, a feature we will explore more in later
chapters.

104 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Chapter 8
Working with Flow and Capacity: The DMV
Almost every object in the SIMIO standard library has capacity, which means that their capacity may be
seized and released, typically within the flow of entities - but the SERVER and the RESOURCE objects are used
the most often. As seen in earlier examples, the SERVER is often the basic flow component of many models.
When the RESOURCE has been employed, it is often a secondary resource to a SERVER. But their relationship
can be convoluted by the need to consider systems with multiple capacities in the context of complicated
entity flow.
The SERVER object is composed of three stations (locations) where entities are held (i.e., INPUTBUFFER,
PROCESSING, and OUTPUTBUFFER). They are illustrated with “contents” queues, each having their own
capacity (see Figure 8.1). A BASICNODE is used as input to the SERVER while a TRANSFERNODE for output.

Processing Station
Input Buffer Station
(Server1.Processing.Contents)
(Server1.InputBufffer.Content
s) Output Buffer Station
(Server1.OutputBufffer.Contents)
Figure 8.1:: The Server Object
The capacity of the SERVER can be a fixed value or have its capacity controlled by a WorkSchedule. Recall
that capacity simply refers to the number of entities that can be simultaneously served by the object. When no
capacity of the processing station is available and an INPUTBUFFER exits (i.e., Input Buffer capacity is greater
than zero), then the entities wait in the INPUTBUFFER. The Processing Time property is the time required for
the server to process an entity in the processing station and is a required property of the server. 94 The SERVER
can occupy one of five states as explained in Figure 8.2 and SIMIO provides additional symbols to
automatically animate the current state of the SERVER. The state applies to all its capacity (i.e., if one unit of
its capacity is allocated, the server is considered processing) and statistics are collected on the amount of time
spent in each state.

Server State Meaning Additional Symbols

Currently idle since no unit of
Starved
capacity has been allocated.
At least one unit of capacity has Server states
Processing
been allocated to process an entity. have a color, a
The current entity has finished name, and a
Blocked processing but cannot proceed into number
the network
The server has failed because of
Failed
some condition

Currently the capacity of the Server

OffShift
is zero meaning it is unavailable.
Figure 8.2: Active Symbol Definitions

94
The Process Type property can be changed to handle more complicate sequences rather just a “Specific Time” and will be explored
in a later chapter.
Working with Flow and Capacity: The DMV 105

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 A RESOURCE is a much simpler object than the SERVER but shares a lot of common properties including
capacity specification but is not composed of stations. Therefore has no place for entities to wait or be served
and there is no processing time specification used to control when the resource is released. So a resource must
have its capacity seized and released in a SIMIO process externally, as seen previously.
Flow and capacity are often the primary concern in many different types of systems (i.e., production,
manufacturing, service, healthcare, etc.). Capacity maybe used to restrict flow but cannot be arbitrarily
expanded. So finding ways to expand capacity and extend flow are significant challenges where simulation
offers one way to experiment with different configurations of capacity and different flow mechanisms.

Part 8.1: The DMV Office

A Department of Motor Vehicle (DMV) often operates multiple offices of various sites across a state. If a
simulation model can be created for one site, then perhaps it can be "reused," with some minor changes, for
other sites (and thus the cost of model development is spread over several installations). A typical DMV
office tests driver license applicants and supplies driver licenses to those passing the tests. A typical office
size is being used as a reference for the analysis. The office typically opens up at 8:30am and closes at 4:00
pm each day to service license applicants. Applicants arrive more or less randomly although the rates of arrival
changes during the day. There are three types of applicants for licenses: Permit, New License, and Renewal.
All arriving applicants go through registration where one of the DMV clerks enters an applicant’s personal
information into the computer and collects the license fee. Next all applicants take a written and visual test on
one of three test stations, each of which are staffed by its own DMV Tester. Not everyone passes these tests.
Those getting a “New License” must also pass a driving test accompanied by a DMV officer where some fail
the driving test. People who fail any (written/visual and driving) of the tests leave the office and return another
day. Those who pass all their tests have their picture taken and their license processed and printed. The DMV
is interested in the service of the office to applicants and the use of DMV resources. Consequently flow times
of applicants are a major concern. Since queuing times are a part of flow time, it is important. If congestion in
the office becomes high, then it also is a concern. Finally, since these are public facilities, the cost of the
facilities, especially the use of DMV resources must be balanced against the service to applicants.
Developing a Base Model
In any modeling activity, try to start with the simplest model you can develop quickly. Make whatever
assumptions you need to get a model “up and running” right away. It’s alright if your assumptions don’t
correspond to reality. We want something working, because we want to extend and evolve our model, rather
than creating a final model at the beginning. Most people new to simulation modeling want their model to be
close to the final one, even though it is their initial attempt. What they discover is that their model has all
kinds of complexities that they have trouble unraveling when they encounter an error. If they started with a
simple (maybe completely unrealistic) model that is working, they can embellish it incrementally and adapt it
into a final working model with far less difficulty than trying to create the final model at the beginning.
Finally, it is not possible to build a model that completely represents the real system in every detail. If that is
your plan, then you are better off experimenting with the real system. Remember a model is only an
approximation of the real system. How detailed your model is depends on what you want the model to reveal.
In general, your model needs only to be as detailed as needed to obtain the performance measures of interest.
Any more detail is unnecessary.95
A second recommendation is not to worry about the input right away. The critical concern is to get a model
that appears most closely to match the system, so you can obtain the needed performance measures reliably.
As you are building your model, it will become clear what input data is needed. Let the model inform your
data collection, so you don’t spend a lot of time collecting the unimportant or wrong data.

95
A problem often arises when a client/boss sees the animated model not conforming exactly like the real system. Some people
(naively) equate the “quality” of the model with the quality of the animation. Sometimes you are forced into animation details that
have little to do with performance measures.
106 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 1: Draw a flowchart of the service process which can come from a “value-stream map”.

Two Testers
0m Fail Driving
Driving Test
Test
20 m
20 m
Two Stations
Applicants 5m 5m Written/ 5m Picture/ 5m Applicants
Registration
Arrive Visual Tests License Leave

Three Stations Two Photo Stations

10 m Fail Written/
Visual Test

Figure 8.3: DMV Flowchart

Step 2: Since our flowchart in Figure 8.3 corresponds easily to the basic SIMIO objects, it can be used as a
basis for the initial model. Create a model with the appropriate objects as seen in Figure 8.4.

Connector

BasicNode

Figure 8.4: Initial DMV Model

· Insert a new MODELENTITY named EntApplicants with the Initial Desired Speed property set to “1”
meter per second. The applicants arrive randomly, but in a time-varying arrival process Based on
information, general estimates of the arrival rates are shown in Table 8.1. From the Data tab, insert a
new RATE TABLE named ArrivalRateTable that has 15 intervals with an interval size of 30 minutes.
Table 8.1: Hourly Arrival Rates for Applicants
Time Hourly Arrival Rate
8:30-9:00am 15
9:00-9:30am 10
9:30-10am 8
10:00-10:30am 12
10:30-11:00am 15
11:00-11:30 am 18
11:30-Noon 15
Noon-12:30pm 12
12:30-1:00pm 10
1:00-1:30pm 8
1:30-2:00pm 6
2:00-2:30pm 7
2:30-3:00pm 5
3:00-3:30 pm 5
3:30-4:00pm 5
· Insert a new SOURCE named SrcArrive that uses the new arrival table ArrivalRateTable.

Working with Flow and Capacity: The DMV 107

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 · The four basic activities will be modeled using SERVERS named SrvRegistration, SrvTests,
SrvDrivingTest and SrvPicture as seen in Figure 8.4.
· From Figure 8.4, notice the BASICNODE named BNodePass that was inserted after SrvTests which is
not part of the flow chart but will be used to help branch applicants who fail the written or eye test.
Use a CONNECTOR between the SrcTests SERVER and the BASICNODE so it doesn’t take any time. Use a
20 and a five meter path between the BASICNODE and SrvDrivingTest and SrvPicture respectively.
Connect all other objects via paths based on the distances from the flow chart.
· Finally, insert three SINKS named SnkLeave, SnkFailTest, SnkFailDriving.96

Step 3: Recall there are three types of applicants (i.e., permits, new licenses, and renewal licenses).

· In order to keep track of time in system for each type, insert a TALLY Statistic as seen in Figure 8.5.
Make sure your TALLY statistics have Time as the Unit Type.

Figure 8.5: Creating Tally Statistics to Track Time in System for Each Type of Applicant
· From the Data tab, insert a new data table named TableApplicant that has six different properties
with the following particular characteristics defined in Table 8.2 and seen in Figure 8.6.
Table 8.2: Applicant Characteristics Data Table
Property Name Property Type Description
Name String The name of applicant type
ApplicantType Integer Will be used to set the entity’s animation picture
ApplicantMix Real The percentage of time this applicant type arrives
Expression (Unit Type set
RegistrationTime to “Time” with minutes The registration time varies by the applicant type
as the Default Units).
PassTests Real Probability of passing varies by applicant type
PassDrivingTest Real Probability of passing the driving test
TallyStatTimeInSystem Tally Statistics Element Used to track each type’s time in system

Figure 8.6: Table Values for the Three Applicant Types.

· Select the EntApplicant and add two additional symbols coloring them red and blue respectively.97

96
Note, different SINKS will be used to separate the statistics for those that fail test versus not. From an animation standpoint, one can
put the three SINKS on top of one another to demonstrate the applicants leaving the same door so it visually looks correct.
97
Switch to 3-D mode to color the entire entity widget.
108 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 · Next, as EntApplicants are created by the SrcArrive, assign the applicant type and associated picture
using the Created Entity Add-on Process Trigger as seen in Figure 8.7.98

Figure 8.7: Assigning the Applicant Type and Picture

· Once applicants leave the system, tally the time in system through the Entered Add-on Process
Trigger of the SnkLeave as shown in Figure 8.8.

Figure 8.8: Tally Time in System

Question 1: Will these Tally statistics by applicant type include the time in system for those applicants that
fail any of the tests (why or why not)?
________________________________________________________________________________

Step 4: Use Table 8.3 to set the processing times and capacities for each of the four servers.
Table 8.3: Processing Time Property Values
SERVER Processing Time Property Value Initial Capacity
SrvRegistration TableApplicant.RegistationTime 2
SrvTests Pert(10,14,20) minutes 3
SrvDrivingTest Pert(15,25,40) minutes 2
SrvPicture Pert(1,2,3) minutes 2

Question 2: Why would the Pert distribution be preferred over the Triangular for Processing Time?
________________________________________________________________________________

Step 5: If you save and run the model, the applicants will choose the various branches equally likely since
the Selection Weight properties are all set to one. Recall only new licenses go through the driving test and
the passing rate of the driving and written/eye tests are stored in the table. Change the Selection Weight
properties for the paths so they account for how the applicant types should branch as seen in Table 8.4.

98
We have a strong preference for processes over the specifications in objects (e.g., the table, state, tally properties), since processes
give use the most flexibility in specifying and re-specifying these additions to the model.
Working with Flow and Capacity: The DMV 109

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Table 8.4: Branching Conditions
From Node To Node Selection Weight Property
BASICNODE named
Output@SrvTests TableApplicant.PassTests
BNodePass
Output@SrvTests 100 – TableApplicant.PassTests
Input@SnkFailTests
TableApplicant.ApplicantTypeID ==
BNodePass Input@SrvDrivingTest 1
TableApplicant.ApplicantTypeID !=
BNodePass Input@SrvPicture 1
Output@SrvDrivingTest Input@SrvPicture TableApplicant.PassDrivingTest
100 –
Output@SrvDrivingTest Input@SnkFailDriving
TableApplicant.PassDrivingTest

Step 6: Save and run the model, observing it for a while before fast-forwarding to complete one
replication.

Question 3: What are the average times in system for New License, Permit, and Renewal applicants?
________________________________________________________________________________

Question 4: What are the observed “ScheduledUtilization” of the stations for Driving Test, Picture,
Registration, and Tests?
________________________________________________________________________________

Question 5: Do you have any concerns relative to the behavior of the model or the results?
________________________________________________________________________________

Question 6: Why do we need to be careful about drawing conclusions from a simulation composed of only
one replication?
________________________________________________________________________________

Question 7: How many units of capacity need to be unavailable for the resource to occupy the
“Offshift(4)” state?
________________________________________________________________________________

Question 8: What bothers you about the model realism, as seen in its animation, based on your experience
in a DMV office?
________________________________________________________________________________

Part 8.2: Using Resources with Servers

After building the initial model, it was discovered that the people who register applicants also take the pictures
as well as hand out the licenses. These DMV clerks work in the same physical area and share the workload,
meaning there are four clerks to perform both functions. However, when we ask about the working
environment, we find that only three applicants can be processed at once at the registration and that only two
people can have their picture taken at the same time.

Step 1: Change the capacity of the SrvRegistration to three, meaning that no more than three entities can
be processed at the same time at this server. The SrvPicture’s capacity will remain two.

Step 2: Insert a RESOURCE object into the model named ResClerks with an initial capacity of four which
corresponds to the four clerks. A resource object can have its capacity “used” at several servers within a
model while server capacity is limited to a specific server.

110 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 9: Even though there are four clerks, why can only three be used at the registration process at the
same time?
________________________________________________________________________________

Step 3: Next, at both the SrvRegistration and SrvPicture servers, use their Secondary Resources99
section for “Resource for Processing” and specify the use of ResClerks100 as seen in Figure 8.9.

Figure 8.9: Specifying the use of Clerks at the Server

Step 4: Save and run the model.

Question 10: Does the model now behave as expected?

________________________________________________________________________________

Question 11: Did you notice that the RESOURCE shows its “Busy (1)” state whenever any of its capacity is
being used similar to the Processing state of the server?
________________________________________________________________________________

If a server employs resources for service, then the capacity to serve entities may be limited by either the
resource capacity availability, the server capacity availability, or both. If only resources limit the capacity,
then the server capacity should be made arbitrarily large (i.e., infinity) so as to not constrain the system.

Part 8.3: Handling Failures

Suppose we now discover that the reception area is interrupted by phone calls. These phone calls arrive
randomly (i.e., Exponentially distributed) on average every 30 minutes and they take anywhere from 1 to 4
minutes, with 2 minutes being typical.

Both SERVERS and RESOURCES can experience failures as can WORKSTATIONS, COMBINERS, SEPARATORS,
VEHICLES, and WORKERS. The failure options are found under the object’s “Reliability Logic” property section.
In SIMIO, the failure concept is intended to be used for any kind of downtime, exception, or (non-modeled)
interruption101. There are four different types of failures that can be specified, two of which are based on
“counts” while the others are based on time as described in Table 8.5.

99It is convenient to use the Secondary Resources when you don’t need the flexibility of the add-on processes. Resources for
Processing does both the seizing and releasing of the resources.
100
By default the capacity requested will only be one. You have more flexibility with the Seize step.
101 An Interrupt step can be used to model interrupts.
Working with Flow and Capacity: The DMV 111

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Table 8.5: Different Types of Failure Types

Failure Type Description

No Failures No failure or exception will occur.
The user specifies the time between exceptions or failures. Often used model general
Calendar Time Based
reliability or maintenance programs.
The user specifies that an exception or failure will occur after a particular number of events
Event Count Based
has happened (e.g., the number of tool changes or certain number of arrivals).
Processing Count The user specifies that an exception or failure will occur after a certain number of operations
Based has occurred the SERVER.
The user specifies that an exception or failure will occur after the object has operated a
Processing Time certain amount of time (i.e., after a 100 hours of operation the machine has to go through
Based maintenance).

When an object is interrupted, its entire capacity is interrupted – its state is changed to “Failed(3)”. So do we
interrupt the service at the server or do we interrupt the resource?

Step 1: To model the phone call interruption, one can use a “Calendar Time Based” type failure. Select the
SrvRegistration SERVER object and specify the “reliability logic” as shown in Figure 8.10, using the
Uptime Between Failure and the Time to Repair expressions corresponding to the phone calls.

Figure 8.10: Specifying the Failure

Step 2: Implement this reliability logic in the SrvRegistration. Run the model and observe its behavior.

Question 12: Did you see the interruptions, displayed by the sever being colored red?
________________________________________________________________________________

Question 13: If you simulate in fast-forward the simulation ends at 7.5 hours, how many phone calls were
handled at the registration desk and what percent of time was the registration desk failed?
________________________________________________________________________________

Question 14: Did the failures of the server cause the secondary resource to fail? What was its state while
the server is failed?
________________________________________________________________________________

Question 15: Did the interruptions in the server affect the “TimeProcessing” in the registration?
________________________________________________________________________________

Step 3: In reality, the clerks are the ones that need to answer the phone calls. Take the failure off the server
and put it on the resource ResClerks. Run the model and observe its behavior.

Question 16: Did you see the interruptions, displayed by the resource being colored red?
________________________________________________________________________________

Step 4: If you simulate in fast-forward the simulation ends in 7.5 hours. Examine the resource statistics for
ResClerks and you will notice a “TimeFailed” and a “TimeFailedBusy” state statistics. The “Failed(3)”
state occurs when the resource is not busy and the “FailedBusy(5)” state occurs when it is busy (at least
112 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 one unit of capacity is being used). Notice that the colors for both states are the same, so you can’t tell in
the animation between the two failed states.

Question 17: From the statistics on the resource, how many phone calls were handled by the resource when
it was idle and how many phone calls were handled when the resource is busy?
________________________________________________________________________________

Question 18: Did the interruptions in the resource affect the “TimeProcessing” in the registration?
________________________________________________________________________________

Question 19: Why do the interruptions of resource cause an increase in the time being processed?
________________________________________________________________________________

Step 5: Unfortunately, failing the entire server or resource is probably not our best option for modeling the
handling of phone calls. Since the phone calls are answered individually, we need to spread them over the
clerks. One possibility would be to create entities to represent the phone calls and have them interrupt
existing service. However, for our present purposes, let’s use the pre-defined reliability logic. However
each clerk needs to handle phone calls individually which means we need to have individual resources.
Replace our ResClerks with four separate RESOURCE objects, each with capacity one (see Figure 8.11).
Since phone calls occur about 30 minutes apart (on the average), we will specify the time between phone
calls to be 120 minutes for each clerk, using the previous time to repair.102

Figure 8.11: Individual Clerks

Step 6: Unfortunately, secondary resources103 can only seize a single specific object (i.e., the ResClerks
which had a capacity of four) but now there are four separate RESOURCES that need to be seized. From the
“Definitions” tab, create an OBJECT LIST called ListClerks that contains the four clerk Resources as
seen in Figure 8.12. The list will allow us to select one of the clerks based on some criteria.

Figure 8.12: Creating an Object List of Resources

Step 7: At both the SrvRegistration and SrvPicture SERVERS, under the Secondary Resources section for
Resource for Processing choose the “Select From List” option as the Object Type property. Specify the
ListClerks as the Object List Name property value as seen in Figure 8.13. Using “Random” as the
Selection Goal forces SIMIO to uniformly use the clerks at the servers.

102
You can modify the existing resource object so it has capacity one and its Uptime Between Failures as Random.Exponential(120)
minutes. Then copy the object three times and name them accordingly.
103
The Seize process step has the same issue of being able to seize from a single object.
Working with Flow and Capacity: The DMV 113

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 8.13: Select a Clerk
Step 8: Save and run the model, both looking at the animation and finally looking at the statistics at the end
of 7.5 hours. Now four resources are being modeled, so there are individual statistics for each.

Question 20: What is the ScheduledUtilization for each of the clerks?

________________________________________________________________________________

Question 21: How many phone calls did each of the clerks process while they were idle?
________________________________________________________________________________

Question 22: How many phone calls did each clerk process while they were busy?
________________________________________________________________________________

Part 8.4: Server Configuration Alternatives

Sometimes there are alternative desirable configurations for the servers in a model, often motivated by
animation needs or handling separate failures. For example, the SrvTests object in the current model
represents the processing of up to three applicants at the testing stations. Perhaps it would be better to
represent this service as three separate SERVERS. Also, the evaluators that are doing the test take a break after
processing between eight and twelve applicants, so separate SERVERS will be needed.

Step 1: The evaluators that assess the written and eye tests take a break after processing between eight and
twelve applicants. The reliability logic can be used to handle this type of break. Set the Failure Type
property to “Processing Count Based” to count the number of applicants where the Count Between
Failures set to a uniform distribution between eight and twelve with the time of the break taking between
two and six minutes but generally they take four minutes as seen in Figure 8.14.

Figure 8.14: Taking a Break Based on the Number Applicants Processed

Step 2: Since the evaluators doing the test will not break at the same time, we need to have individual test
stations. Change the name of SrvTests to SrvTest2 and the capacity back to “1.” Copy it two times
creating a total of three SERVER objects as shown in Figure 8.15. Connect the test stations with the
registration using five meter PATHS. We will assume that the branching to each testing station is equally
likely (i.e., equal Selection Weights). Connect the other two tests with CONNECTORS to the Ouput@
SrvTest2 which allows us to use the same branching logic to driving test, picture taking, or failure.

114 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 5 Meter Path

Connector

Figure 8.15: Separate Testing Stations

Step 3: Now save and run this model observing the animation.

Question 23: How do the applicants choose which of the three testing stations to enter?
________________________________________________________________________________

Question 24: Is there anything about the behavior of this model that causes you concern?
________________________________________________________________________________

Step 4: Typically there is one waiting station after registration where they wait for the next available
testing station. To prevent the waiting of applicants at the testing stations, first set the input buffer capacity
of each of testing servers to zero under the Buffer Capacities section of the SERVER object.

Step 5: Save and rerun the model, observing the animation.

Question 25: Now, how do the applicants wait at the testing stations? Is that a desired behavior?
________________________________________________________________________________

The applicants now wait at the input node of the testing stations, unable to enter the SERVER’S station since
the input buffer capacity is zero. Furthermore the entities seem to make a decision about which path to take
independent of the traffic on that path, the number at the associated test station, or if the tester is on break.

Step 6: To handle those issues, the branching process needs to be modified to choose an appropriate testing
station. From the “Definitions” tab, create a NODE LIST called ListTesters that contains the input node
(i.e., Input@) for each of the three testers as seen in Figure 8.16. Having a list will allow SIMIO to select
one of the input nodes using some criterion.

Figure 8.16: Creating a Node List of Testers

Step 7: Select the Output@SrvRegistration transfer node that connects the registration to the servers,
change the Entity Destination Type property to “Select From List” and in the Node List Name property
select the ListTesters list that was just created.

Working with Flow and Capacity: The DMV 115

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 · Set the Selection Goal property to be the Smallest Value with the Selection Expression104 equal to
Candidate.Node.AssociatedStationload105 which chooses the server with the shortest
line by determining the number that are routing to the SERVER plus the number in the Contents
Queues of the InputBuffer106 and Processing107 stations as seen in Figure 8.17.
· If a tester is taking a break, applicants should not be routed to this person while they are on the break.
To avoid selecting these nodes, the Select Condition property can be used to include only certain items
from a list. This expression property has to evaluate to “True” before the CANDIDATE NODE object will
be considered for selection for routing. Therefore, we will set the property to select only those servers
which have not failed before choosing the smallest line
· The Failure.Active function will be “False” if the server is currently not in a failure mode. If
the server has not failed, then allow it be a candidate to be selected using the expression
!Candidate.Server.Failure.Active as seen in Figure 8.17.108

Figure 8.17: Choosing the Shortest Line

Step 8: Save and run the model observing what happens to the entities that leave registration. You may
need to increase the time of the breaks to make sure the selection condition is working.

Question 26: Now does the model appear to be behaving as you want?
________________________________________________________________________________

Part 8.5: Restricting Entity to Flow to Parallel Servers

Now applicants are forced to wait in the registration output node until a tester is available (i.e., the SERVER is
empty) owing to the traveler capacity paths being one and the input buffer of the servers are zero. In a later
chapter we will present a more complete explanation of various blocking methods. Sometimes it is necessary
to block flow under somewhat complex circumstances. Suppose that the testers, who give the visual and
written exam must enter the results of the exam into a computer. The applicant is allowed to return to the
waiting area, but the tester remains busy to complete the process. Obviously, the tester is unavailable to other
applicants until the results are entered.

104 Do not confuse the Selection Expression property with the Selection Condition which is used to include/filter candidate nodes.
105
Note the CANDIDATE object has to be used to specify a wildcard like ‘*’. SIMIO will replace the CANDIDATE object with the
appropriate object from the list when determining the smallest value. Also, if there is a tie between nodes they will be selected in the
order they were placed into the list.
106
The Contents queue of the InputBuffer Station is used to hold entities waiting to be processed by the SERVER.
107
The Contents queue of the Processing Station represents the location of the entities that are currently being processed by
the SERVER.
108
The ! or Not operator can be used to take the complement of a variable (i.e., ! False would equal True).
116 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 1: To keep track of the status of the tester, we will utilize a STATE VARIABLE associated with each
tester. From the Definitions→States section, insert a new discrete String state variable named
GStaTesterStatus. Since there are three rooms, make the Dimension Type property be a Vector and set
the number of Rows too three with an Initial State Value of Available109 as seen in Figure 8.18.110

Figure 8.18: Setting a Vector State Variable

Step 2: From the Animation tab, insert three status labels beside each of the tester SERVERS with an
expression associated with the GStaTesterStatus (e.g., GStaTesterStatus[1] for the first room as
seen in Figure 8.19). Specify all three expressions accordingly. Vectors start with an index of one and each
row is specified in brackets.

Status Label

Figure 8.19: Adding Status Labels

Step 3: Once an applicant has been seen by one of the tester, we will need to set the status of the tester to
“DataEntry”, delay for the five to ten minutes, and then set the status to “Available”. We will model this
via a specialized process for each room. Navigate to the “Processes” Tab, from the Process section, use
the Create Process button to create a new process named ProcessDataEntry, as seen in Figure 8.20.
Note, this is a case where State Assignments cannot be used to model the scenario.

Figure 8.20: Process for Data Entry and Changing the Status of the Tester
· Insert an Assign step that will set the first room’s status (i.e., GStaTesterStatus[1]) to
“DataEntry”.

109
Here “Available” is a string constant (arbitrarily chosen), not a numerical value.
110
Note we could have created three individual state variables (i.e., GStaTesterStatus1, GStaTesterStatus2, and GStaTesterStatus3)
instead of the one variable that is a vector of size three.
Working with Flow and Capacity: The DMV 117

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 · Insert a Delay step that with a Delay Time expression set to Random.Uniform(3,6). For a
moment we will leave the Units set to hours which allow us to see any potential problems.
· Copy the first Assign step and change the Value to “Available”.

Figure 8.21: Process Steps and Values for the ProcessDataEntry

Step 4: The new process needs to be invoked after the applicant has finished processing. In the Facility
window, select SrvTests1 and specify the ProcessDataEntry1 as the After Processing add-on process
trigger which is invoked immediately after the entity has finished processing.

Figure 8.22: Using the ProcessDataEntry1 as the After Processing Add-On Process Trigger
Step 5: Save and run the model.

Question 27: What happens to the applicant that finishes service and then causes the data entry to happen?
________________________________________________________________________________

Step 6: You should have noticed that the entity does not leave the tester until after the data entry completes
which is not what we wanted. The reason this occurs is the MODELENTITY is delayed and not just the tester.
The Exited add-on process trigger is invoked after the entity has physically left the SERVER. Instead of
using the After Processing, specify the ProcessDataEntry process as the Exited add-on process trigger.

Step 7: Save and run the model.

Question 28: What happens to the applicant that finishes service now and causes the data entry to happen?
________________________________________________________________________________

Question 29: What happens to the next applicant waiting for this tester?
________________________________________________________________________________

In the first case, the current entity was delayed and did not leave the tester. When using the Exited trigger, the
current applicant leaves the room but the next applicant is not blocked, will enter the tester, and begin
processing. Also, the utilization of the tester is not correct because they are idle when performing the data
entry.

Step 8: Therefore, we need to seize the tester to prevent it from being freed when the applicant leaves the
room and release the server once the data entry is finished. Insert a Seize and Release steps to the
ProcessDataEntry process as shown Figure 8.23.

118 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 8.23: Adding a Seize and Release of Tester1
· Because the entity has left the server, it cannot be used to seize the tester. Therefore, we will use the
ParentObject which is the MODEL in this scenario as the Owner Type for both the Seize and
Release process steps as seen in Figure 8.24.

Common to both Seize

and Release
Figure 8.24: Seize and Release SrvTest1 by ParentObject
Step 9: Save and run the model observing whether the SrvTest1 server does not accept entities now.

Question 30: Does the SERVER accept entities while doing data entry?
________________________________________________________________________________

Step 10: The entities are still allowed to enter the path to the server even though the server is busy.
Therefore, we need to block the server or the path using a different mechanism.111 Recall, the Selection
Condition property of the Routing Logic section has to evaluate to “True” in order for that particular node
to be selected. Currently the condition will not select testers that have failed. Therefore, modify the
condition to include not choosing servers that are currently processing (i.e., resource state equal to one) as
seen in Figure 8.25.
!Candidate.Server.Failure.Active && Candidate.Server.ResourceState != 1112

Figure 8.25: Restrict Testers that are Busy from Being Selected
Step 11: Save and run to make sure that the SERVER does not accept entities now.

111 One could shut down the path by changing the capacity or the direction to None or potentially cause the TransferNode Current

Travel Capacity state variable to go to zero.

112 Note, the && is for logical “And” such that we only select servers that have not failed and currently not processing someone.
Working with Flow and Capacity: The DMV 119

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 31: So if the server’s resource state is one (i.e., “Processing) when doing data entry, how
does this expression insure that only idle servers are eligible for selection?
________________________________________________________________________________

Question 32: Does the SERVER SrvTests1 accept entities while doing data entry now?
________________________________________________________________________________

Step 12: Since the model is working properly now, change the Units of time to “Minutes” for the Delay
step of ProcessDataEntry.

Step 13: At this point, we could select the ProcessDataEntry process and copy it two times. Then change
the names of the two copied processes to ProcessDataEntry2 and ProcessDataEntry3, the two Assign
steps (i.e., the Row property set to two or three), the Seize step, and the Release step to specify the
correct tester. However, to do this more generally (i.e., have the same process used by all three testers),
we will use a “Tokenized Process.” This approach will be especially useful if additional testing stations
are added later.

· From the Definitions tab, define a new TOKEN named TknTester. Add an INTEGER STATE variable
name TStaWhichRowNumber and from the Object Reference section add an OBJECT REFERENCE
STATE variable named TStaWhichTester as seen in Figure 8.26.

Figure 8.26: New Token TnkTester to be used in a Tokenized Process

· Modify the ProcessDataEntry1 process by first changing its name to ProcessDataEntry. Next
modify the process’s Advanced Options properties by specifying the process will be executed by the
TknTester TOKEN as shown in Figure 8.27. Next, specify the process will take two input arguments
(i.e., WhichRowNumber linked to TknTester.TStaWhichRowNumber and WhichTester linked to
TnkTexter.TStaWhichTester).

Argument Passed

Figure 8.27: Adding Token Information to the Process

Step 14: Next change the two Assign steps within the process to employ the TStaWhichRowNumber as
the Row property to update the GStaTesterStatus passed to the TOKEN TknTester as seen in Figure 8.28.

120 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 8.28: Use the Token Argument
Step 15: Update the Seize and Release steps to use TStaWhichTester as the specific Object Name.

Figure 8.29: Using the TknTester to Specify Which Server to Seize and Release
Step 16: From the facility window, select SrvTests1 and set the Exited add-on process trigger as seen in
Figure 8.33 to pass in the correct row number and tester. Repeat for the other two testers setting the row
number and which tester to the correct values (i.e., “2” and “3” and SrvTests2 and SrvTests3
respectively).

Values Passed

Figure 8.30: Invoking the Exited Process with its Argument

Step 17: The status labels currently change from “Available” to “Data Entry” but remain “Available” while
the SERVER is processing an applicant. To fix this issue, we need to update the status to busy when the
tester is servicing an applicant.

· Since it will be similar for all three testers, add another “Tokenized Process” named
ChangeStatustoBusy as seen in Figure 8.31. Next use the Token TknTester as the Token Class
Name property and insert one input argument (i.e., the row number).

Figure 8.31: Creating the Process that will update the Status to Busy
Working with Flow and Capacity: The DMV 121

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 · Insert an Assign step that will set the GStaTesterStatus variable to “Busy” as seen in Figure 8.32.

Figure 8.32: Setting the Status to Busy in the Assign Step

Step 18: For each of the three tester SERVERS, specify the Before Processing Add-on Process triggers to use
the ChangeStatustoBusy process specifying the correct row number as seen in Figure 8.33 for the first
tester.

Figure 8.33: Changing the Room Status to Busy

Step 19: Save and run the model making sure the status labels turn to “Busy” and that the model is working
properly.

Question 33: What is the average utilization of each server?

________________________________________________________________________________

Question 34: What is the total time an applicant has to wait for a tester?
________________________________________________________________________________

Question 35: What is the total cycle time for each license type?
________________________________________________________________________________

Question 36: Are there any drawbacks of this new model?

________________________________________________________________________________

Part 8.6: The Waiting Room Size

In the real system people wait for services in a common waiting area, except for those waiting for registration
which queue in front of the registration. This common area consists of waiting for the tester, waiting for the
driving test, and waiting for the picture/license. Looking at each waiting area separately ignores the fact that
contributions to the waiting area changes throughout the day. To visually see what is happening in the
common waiting area, we need to introduce the SIMIO STORAGE object which will be used to represent all the
waiting areas together. (Storages do not actually contain entities but entities hold membership in storages.)

Step 1: From the Definitions→Elements section, insert a new “General” STORAGE element named
StoWaitingRoom.

Step 2: Entries (i.e., entity memberships) are added to STORAGE elements via an Insert step and removed
via the Remove step. Add two new processes named WaitingRoomEntered and WaitingRoomExite as
seen in Figure 8.34 and Figure 8.35 respectively. Specify the StoWaitingRoom.Queue as the Queue
State Name property.

122 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 8.34: Inserting into the WaitingRoom

Figure 8.35: Removing from the WaitingRoom

Step 3: People enter the common waiting room when they leave the registration desk or when they are
waiting at the driving test or the picture area.
· For the After Processing add-on process trigger of the SrvRegistration specify the
WaitingRoomEntered process.
· Select both the SrvDrivingTest and SrvPicture and then specify the WaitingRoomEntered process
as the Entered add-on process trigger.

Step 4: People leave the waiting room when they enter service at the testers, driving test or picture area.
· Select both the SrvDrivingTest and SrvPicture and then specify the WaitingRoomExited process as
the Before Processing add-on process trigger.
· For the three testers (i.e., SrvTests1, SrvTests2, and SrvTests3), we will need to modify the
ChangeStatustoBusy process since it is already specified as the Before Processing add-on process
trigger. Insert an Execute step that with the Process Name specified as WaitingRoomExited.

Figure 8.36: Using Execute Step to Cause Applicants to Leave the Common Area
Step 5: To “see” the waiting room, insert a DetachedQueue from the Animation tab. The Queue State
property for the queue is the StoWaitingRoom.Queue. Draw a straight line for the queue.

Step 6: Most waiting rooms have people along the edge of a room rather than a straight line.113

· Select the animated WaitingRoom queue which by default is the green straight line.
· Change the Alignment to Oriented Point and then select the Keep in Place option since entities once
they sit will not keep changing seats along the queue like a checkout line does as seen in Figure 8.37.
· From the Vertices section, add two vertices by selecting the Add Vertex button and dragging the green
vertex ( ) onto the queue.

113
From an animation standpoint, we would delete the input buffer queues of the testing stations, driving test, and picture area.
Working with Flow and Capacity: The DMV 123

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 8.37: Adding an Animated Queue
· Form a U-shaped queue by moving the two new vertex points that were just added to the appropriate
spots. Then continue adding vertexes and then moving the points inward to form the waiting room
queue as seen in Figure 8.38.

Use as Waiting
Room

Figure 8.38: Adding an Animated Queue in a Different Shape

Step 7: Save and Run the model.

Question 37: What is the average number and the maximum number in the waiting room (storage)?
________________________________________________________________________________

Question 38: How much time, on the average, is spent in that waiting area?
________________________________________________________________________________

Part 8.7: Using Appointment Schedules

In response to applicant waiting concerns, the office has decided to offer a few scheduled appointments to
provide better service to the applicants. The office will still allow most people to “walk-in” but will give
priority to those who have an appointment. In particular, the office will schedule applicants for the following
nine appointment times during the day: 8:30 am, 9:00 am, 9:30 am, 10:00 am, 2:00 pm, 2:30 pm, 3:00 pm,
and 3:30 pm. The appointments schedule applicants during the times when there are lower arrival rates. Some
applicants who have appointments will arrive early while others may arrive late which are called “early/late”
arrivals. Also, it is expected that some may skip their appointment all together (i.e., “no-shows”).

You need to realize that a “scheduled arrival process” is fundamentally different from the usual “random”
arrival process in that the random arrival process has random interarrival times while the scheduled arrival
process has known interarrival times. Randomness in the scheduled arrival process occurs in the early/late
arrivals and the no-shows. In many “service industries” scheduled arrival processes are more prevalent than
random arrival processes.

Step 1: The scheduled appointment times are specified via a data table. From the Data tab, add a new data
table named TableArrivals. Insert a Date Time column named ScheduledTime from the Standard
Property drop down, as seen in Figure 8.39. Next specify the scheduled appointments which indicate just
the time of arrival.114 At this point, these appointments are absolute arrival times (i.e., if the simulation

114
Note, you can build the table in Excel as it can be easier and then copy the table into SIMIO.
124 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 was to start on 9/5/2015 instead of 9/6/2015 no arrivals would occur until the next day or if the simulation
start date was 9/7/2015 then no arrivals would occur). See the Commentary section at the end of the
chapter on specifying a table with relative time offsets.

Figure 8.39: Specifying the Appointments

Step 2: Since we are assuming one person will arrive at the scheduled appointments, we will reduce the
arrival rate table by one for the first and last four time periods so comparisons can be made as seen in the
partial table.
Table 8.6: Hourly Arrival Rates for Applicants
Time Hourly Arrival
Rate
8:30-9:00am 14
9:00-9:30am 9
9:30-10:00am 7
10:00-10:30am 11
Middle Times Unaffected
2:00-2:30pm 6
2:30-3:00pm 4
3:00-3:30 pm 4
3:30-4:00pm 4

Step 3: From the Facility tab, insert a new SOURCE object named SrcScheduledArrivals as seen in Figure
8.40. Next, connect it to the output node of the original SOURCE via a CONNECTOR which allows arrivals to
originate from one point and travel along the same five meter path to the reception area.

Connector

Figure 8.40: Adding an additional Source to Handle the Scheduled Appointments

Working with Flow and Capacity: The DMV 125

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 4: To utilize appointment schedules, change the Arrival Mode to “Arrival Table” and specify the
Arrival Time Property to be the column TableArrivals.ScheduledTime as seen in Figure 8.41.

Figure 8.41: Specifying the Appointment Scheduling in the SOURCE

Step 5: The Arrival Time Deviation property is used to model the deviation from the appointed scheduled
time specified by the TableArrivals.ScheduledTime. In this example, we are stating that applicants can
arrive as early as ten minutes or as late as five minutes but most of them will arrive three minutes early,
using a Pert distribution115. See Table 8.7 for an explanation of all the appointment schedule properties.
Table 8.7: Appointment Scheduled Properties
Appointment Scheduled
Description
Property
The numeric table property that specifies the list of scheduled arrival times
Arrival Time Property
which can be a Date Time or any numeric or expression property.
This is the expected number of arrival events that will occur at each arrival
Arrival Events Per Time
time. You can think of this of how many batches will arrive where the Entities
Slot
Per Arrival will determine the number of Entities per batch that will arrive.
Specifies the deviation from the scheduled time which affects each batch
differently. Therefore, two Arrival Events Per Time Slot and 1 Entities Per
Arrival Time Deviation Arrival will be different than one Arrival Events Per Time Slot and 2 Entities
Per Arrival when a deviation or no-show probability is specified as in the
latter case the two entities will always arrive at the same time.
Arrival No-Show Specify the probability of a no-show occurring which is applied to a batch
Probability arrival.
Entities Per Arrival The number of entities to create for each batch.
When you reach the end of the table, determines whether or not to repeat the
pattern. If you have specified actual times these are converted to deviations or
time offsets. For this example, the first arrival is at 12 PM with the simulation
Repeat Arrival Pattern starting at 11 am which means there is a 1 hour time offset till the first arrival.
So at 4 pm when the last scheduled arrival occurs, the simulation will start
back up with arrivals at 5 pm to have the same offset if this property is set to
“True.”

Step 6: In this case we assume 5% of the applicants are “no-shows” and skip their appointment. Therefore,
set the Arrival No-Show Probability property to 0.05.

115
Negative times mean the entity is early for their appointment, whereas positive number implies they are late.
126 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 7: In the SrcScheduledArrivals source reference the process SrcArrive_CreatedEntity in the
Created Entity add-on process trigger. Also, in order to distinguish the scheduled arrivals assign the
ModelEntity.Priority a value “2” in the Before Exiting of the “State Assignments” section.
Recall the default Priority state variable has a value of 1.116 Refer to Figure 8.42.

Figure 8.42: Specifying Properties of the SrcScheduledArrivals Source

Step 8: We would like to keep track of time in the system according to whether the applicant had a
scheduled appointment or they walked-in. From the Definitions→Elements section, add two TALLY
STATISTICS named TallyStatSchedTimeInSystem and TallyStatWalkInTimeInSystem with the Unit
Type property set to “Time”.

Step 9: To obtain the time in system statistics for the two types of arrivals, each needs to be tallied. One
way to obtain the tallies is to use the Tally Statistics section of the Input@SnkLeave specifying On
Exited as shown in Figure 8.43.

Figure 8.43: Collecting Time in System by Arrival Type

Step 10: Select the experiment and add a new response for each of the TALLY STATISTICS named
WalkInTimeInSystem with an Expression set to TallyStatWalkInTimeInSystem.Average and
ScheduledTimeInSystem with TallyStatSchedTimeInSystem.Average. Both should have a
Unit Type as Time with Display Units as Minutes.

Step 11: Save the current model. Reset the experiment and run for 100 replications.

Question 39: What is the average time in the system and half width for each type?
________________________________________________________________________________

Question 40: Is there any statistical difference in the times in the system for the applicant types?
________________________________________________________________________________

Step 12: One of the potential issues is that scheduled patients are not processed differently in the model.
Applicants who have scheduled appointments should have a higher priority than walk-in applicants.

· Therefore, select SrvRegistration, SrvDrivingTest, and SrvPicture SERVERS in the model and
change their Ranking Rule property so that scheduled appointments have a larger priority as seen in
Figure 8.44.

116
The Priority assignment could be done within the process SrcArrive_CreatedEntity.
Working with Flow and Capacity: The DMV 127

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 8.44: Changing the Ranking Rule in the SERVERS
· For the testing SERVERS, the expression has to be a little more complicated because both entities and
the MODEL seize capacity of server. We need to give priority to the MODEL for the data entry process
using the following expression seen in Figure 8.45. The Is.Entity function will return true if an
applicant is requesting service and false otherwise.

Figure 8.45: Setting Priority Based on Entity and Model

Step 13: Save the model and rerun the experiment.

Question 41: What is the average time in the system and half width for each arrival type?
________________________________________________________________________________

Question 42: Is there an improvement for scheduled appointments?

________________________________________________________________________________

Step 14: There does not seem to be any difference in time when we prioritized the applicants at each of the
servers. To see if the model is working as predicted, let’s add a dynamic label to the entities to display the
priority value. Select the EntApplication MODELENTITY and set the Dynamic Label Text property under
the Animation section to display the priority as seen in Figure 8.46.

Figure 8.46: Adding a Dynamic Label that Moves with the Entity
Step 15: Run the model and observe the output buffer of the SrvRegistration around the 9:30 am time
frame.

Question 43: Are all the scheduled applicants with priority two taken ahead of priority ones?
________________________________________________________________________________

Step 16: The problem is there are three servers which have their own separate sorted queues when we really
want one queue feeding those three servers to be sorted. The issue doesn’t happen at the other areas as the
entities are all in the same queue and sorted properly. To fix the issue, insert a RESOURCE named ResTests
with a capacity of three and Ranking Rule set accordingly to act as this single queue. Also, select the
ResTests and add the ALLOCATION QUEUE from the Draw Queue in the Attached Animation section.

128 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Allocation Queue

Figure 8.47: Adding a Resource to Act as Single Queue Feeding Three Servers
Step 17: Next, we need to seize the new ResTests once the applicants enter the output node of the
SrvRegistration as they try to get service at testing. Select the output node of SrvRegistration (i.e.,
Output@SrvRegistration) and add an Entered Add-On Process Trigger as seen in Figure 8.48. The
ResTests will determine which applicant to serve next and allow the applicant to move on to the testers.

Figure 8.48: Seizing the ResTests in the Output Node of the SrvRegistration
Step 18: Once the entity finishes the testing, we need to release the RestTests. Insert a Release step in the
ProcessDataEntry right after the status is updated as seen in Figure 8.49.

Figure 8.49: Releasing the ResTests After Finishing Testing

Step 19: Save and run the model observing whether or not the applicants are being ranked correctly now.

Step 20: Next, rerun the experiment.

Question 44: What is the average time in the system and half width for each arrival type?
________________________________________________________________________________

Question 45: Is there an improvement for scheduled appointments?

________________________________________________________________________________

Step 21: The applicants are now being processed correctly now. However, on further reflection, if a walk-in
applicant has been in the system for more than 30 minutes then they should be given greater
consideration/priority. Because the time in system changes while the applicant is waiting in the input
buffer, the static Ranking Rule property cannot be used as it orders the queue as the entities enter the
queue and is not reordered later. On the other hand, the Dynamic Selection Rule is evaluated each time the
SERVER capacity becomes available in order to choose the next applicant. Select the SrvRegistration,
SrvDrivingTest, and SrvPicture SERVER objects and specify the “Largest Value First” as the rule as
seen in Figure 8.50 with the Value Expression equal to the following.
(Candidate.Entity.Priority==2)||(Candidate.Entity.TimeInSystem>MaxWaitTIme)

Working with Flow and Capacity: The DMV 129

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 The expression will evaluate to either one (i.e., if the entity is a scheduled applicant or the entity has been
in the system ¾ hour) or zero otherwise.117 Repeat for the ResTests RESOURCE object.

Figure 8.50: Specifying the Selection Rule to Prioritize Scheduled Applicants and Walk-ins
Step 22: Save the model and rerun the experiment.

Question 46: What is the average time in the system and half width for each arrival type?
________________________________________________________________________________

Part 8.8: Controlling the Simulation Replication Length

Of concern is the number of applicants that seem to be in the system at the end of the 7.5 hours that the DMV
is open. After more discussion with the DMV personnel, you realize that the office closes at 4:30 pm, but the
people stay until the last applicant is served. It is easy to run the simulation for exactly 7.5 hours using the
“Run Setup” section of the “Run” tab. But we need the simulation to stop after the last person is served. This
concern raises two modeling issues. First, how do we stop the arrivals after 4:30 pm and second, how do we
stop the simulation run after the last person has left the system? There are a variety of ways to stop the arrival
of applicants after 7.5 hours. For example, we could transfer them to a sink if the current time is after 4:30 pm
(i.e., using selection weights on the paths). Another way would be to create a MONITOR that would cause a
process to alter the interarrival time or the entities per arrival.

Note, the current ArrivalRateTable only has 15 time periods representing the 7.5 hours. If we extend the
simulation run length beyond 7.5 hours, SIMIO will cycle back to the beginning of the table repeating the
values because they are relative offset values. The only way to force zeros is to extend the arrival table to
contain 24 items (i.e., 12 hours) and set the remaining values to zero. A direct method to stop the arrivals is
to use the “Stopping Conditions” within the SOURCE, which turns off arrivals based on either specifying a
maximum number of arrivals118, some maximum time, or some stoppage event.

Step 1: Select the SrcArrive SOURCE, specify the Maximum Time property a value of 7.5 under the
Stopping Conditions section.

Figure 8.51: Stopping Arrivals after 4:30

Step 2: Set the run length to 12 hours and then run model. You will notice that the applicants stop arriving
at 4:30 pm but the simulation continues displaying services (i.e., failures). A simulation run can be

117
If there is a tie (i.e., multiple entities have the same value), they are ordered based on First in First out rule.
118
Note that this is the number of total “arrival events” and not the total number of entities arriving. One arrival event may yield more
than one entity arriving.
130 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 stopped using the End Run step. However this step needs to be invoked under two conditions: (1) at 4:30
pm, in case no one is in the system, and (2) the more likely case when the last applicant is served after
4:30 pm.

Step 3: Let’s first consider the case that the last applicant finishes after 4:30 pm, when no more arrivals are
permitted. Applicants exit the system through one of the three sink objects. So at each of the sinks, we
need to test whether this entity is the last applicant in the system and that it is after 7.5 hours. We will first
create a general process.

· Go to the “Processes” tab and click on the “Create Process” button, naming the new process
TimeToStop. Add a Decide and End Run step as in Figure 8.52.

Figure 8.52: Check Last Applicant and End Run

· The Decide step is “ConditionBased” on the following expression.

(EntApplicant.Population.NumberInSystem <= 1) && (Run.TimeNow >= 7.5)

This expression states that we will stop the replication if the current applicant being destroyed is the last
applicant in the system and the simulation time is greater than or equal to 7.5 hours.

Step 4: Select the “Destroying Entity” add-in process trigger property for each the three SINKS and specify
the TimeToStop process for each.

Step 5: Within the Run→Run Setup arbitrarily set the Run Length to 12 hours, to be sure the simulation
length does not end the run.

Step 6: Save and fast-forward the model to the end of the replication.

Question 47: At what time did the replication stop?

________________________________________________________________________________

Step 7: If you make multiple replications, each replication will stop at a different time. Output statistics
are appropriate for collecting statistics on the time the replications stop. Note, output statistics only collect
one observation for each replication at the end of the run. Such statistics will allow us to determine on
average how long the office stays open. From the Definitions→Elements section, insert a new OUTPUT
STATISTIC named OutStatOfficeCloses with the expression being Run.TimeNow and set the UnitType
to be “Time” and Units to “Hours”.

Step 8: To determine on average of how long the office is open, multiple replications of the model need to
be executed. From the “Project Home” tab, insert a new experiment named ExpOfficeCloses.

Step 9: Change the Required replications for the first scenario to 100.

Step 10: You generally need to add a response to evaluate each of the experiment scenarios. Under the
“Experiment” section, click the Add Response button to insert a response named TimeAfterClosing which
has an expression of the OutStatOfficeCloses.Value - 7.5 which indicates when the

Working with Flow and Capacity: The DMV 131

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 simulation ended minus the 7.5 hours which represents the amount of time over the normal day (see
Figure 8.53).119

Figure 8.53: Specifying a Response for the Experiment

Step 11: Save the project and execute the experiment and from the Pivot grid answer the following
questions.

Question 48: On average how long does the office stay open past the 7.5 hours?
________________________________________________________________________________

Question 49: What is the half width of the time when the office closes (in minutes)?
________________________________________________________________________________

Step 12: Click on the “Response Results” tab, to see the (SIMIO Measure of Risk & Error (SMORE) plot for
the one response as seen in Figure 8.54. SMORE plots allow you to visual the output response. It
concisely displays the minimum, maximum, mean, median, lower and upper percentiles,120 as well as the
confidence intervals of the mean and the lower and upper percentiles. The histogram of values is also
displayed because the “Histogram” button is pushed.

Maximum
Value

Upper Percentile CI

Upper Percentile

Mean Confidence
Interval
Median

Mean

Lower Percentile Lower Percentile CI

Histogram Minimum Value

`
Figure 8.54: SMORE Plot with Histogram for the Additional Time the Office is Open

119
Note we could have made the OUTPUT STATISTIC expression be Run.TimeNow – 7.5 instead.
120
By default it is the 25th and 75th percentiles which can be changed on the Experiment properties along with the default 95%
confidence interval.
132 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 50: Based on the SMORE plot and its “Raw Data”, what are the mean and the 95% confidence
intervals of the additional time office is open?
________________________________________________________________________________

Question 51: What are the lower and upper percentiles and their values?
________________________________________________________________________________

Question 52: What is the maximum additional time spent in the office and does that make sense?
________________________________________________________________________________

It seems that there are occasions that the maximum additional time spent is 7.5 hours (450 minutes) either via
the SMORES plot or the results in the Pivot Grid. The simulation does not stop until the 15 hour run length is
reached. Note, an observation in a sink must occur before the stopping condition is checked.

Question 53: What happens if the last applicant leaves at 3:45 p.m. but no one arrives before 4:30 p.m.?
_____________________________________________________________________________

Step 13: In this situation, the last applicant leaves the system but the current time is still less than the 7.5
hours and therefore the stopping condition is not invoked. The previous stopping method assumes that
there would always be applicants in the system at 4:00 p.m. The model needs to check to see if there zero
applicants in the system at closing. Return to the Model and insert a new TIMER element named
TimerToStop from the Definitions→Elements section and set the Time Interval property to 7.5 hours121.

Step 14: In the Processes tab, insert a new process named StoppingSimulation which utilizes the
TimertoStop event as its Triggering Event Name property and the Triggering Event Condition as seen in
Figure 8.55. When the event occurs the process will only run if the triggering event condition is also
true.

Figure 8.55: Specifying a New Process to be Executed when the Timer fires.
Step 15: Insert the End Run step which will stop the simulation run as soon as the step is executed as
shown in Figure 8.56.

Figure 8.56: Stopping Simulation at Time 7.5 hours Process

Step 16: Save the model and rerun the experiment looking at the “Response Results” again.

121
One timer event will occur at 0.0 since the Time Offset is 0.0, but it won’t affect the simulation in this case.
Working with Flow and Capacity: The DMV 133

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 54: On average how long does the office stay open past the 7.5 hours?
________________________________________________________________________________

Question 55: What is the maximum additional time spent in the office and does that make sense?
________________________________________________________________________________

Question 56: Is it realistic that the office should stay open only 90 minutes past its closing?
________________________________________________________________________________

Question 57: What can be done to reduce the time the office is open after its closing time?
________________________________________________________________________________

Part 8.9: Commentary

· In Figure 8.39, the appointment schedule was designated using absolute dates and times. Internally,
SIMIO converts the table into a time offset table based on the simulation starting date and time. One
can also specify the appointment table directly as time offsets as seen in which represents the exact
same table as the one in Figure 8.57. The column type must be a numeric or an expression general
property.

Figure 8.57: Appointment table Using Time Offsets

134 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Chapter 9
The Workstation Concept: A Kitting Process
One of the objects in SIMIO not yet described is the WORKSTATION which is rather specialized even though it
behaves somewhat like the SERVER object. It was designed to best fit a manufacturing environment, although it
is certainly not restricted to those applications. The WORKSTATION can have scheduled capacity but is limited to
a capacity of only one. The object can mimic batching by processing an entity as though it were composed of
several items, services one or more of the items at a time. There are expanded setup options for the options,
even accounting for dependency between setups. Most of these features are found in actual workstations in a
production system. Perhaps the most interesting feature of the WORKSTATION it is use of “materials”.
Materials are those items which are supplied to workstations that are directly required for
assembly/production. Typically, a “bill of materials” is used to determine what is needed for any particular
operation. Materials are not an object, but an inventory of items which can be consumed and replenished.

Part 9.1: The Kitting Process

A kitting process constructs a “kit” from available components. Orders arrive to a kitting workstation in a
Poisson stream with an interarrival time of three minutes as seen in Figure 9.1. An order is for a particular
number of identical kits to be produced. The kits are assembled from three components, one SubA
subassembly, one SubB subassembly, and two Connectors – this is also referred to as a Bill of Materials
(BOM). The assembly of each kit has a processing time described by a Triangular distribution with
parameters of (0.1, 0.2, 0.3) minutes. Only one kit can be assembled at the kitting workstation at a given time.
Information about the kits produced and components used is needed to analyze the operation, and information
on the processing of the orders is also desired.

Orders Order Quantity:

Arrive Exp(3) min Binomial(0.5,20)+1

Uniform(1,2) min Uniform(1,2) min

30% Green
Assemble Kits
15% Yellow
20% Red Processing: Tria(0.1,0.2,0.3) min
35% Blue Teardown: 1 minute
Setup: 2 minutes

Figure 9.1: The Kitting Operation

There are four types of kits produced from the same BOM, which will be represented by colors: 30% are
green, 15% are yellow, 20% are red, and 35% are blue. Following each order that is processed, it takes one
minute to tear down for the next kitting order and two minutes to setup for the next order. This assumption
will be changed later. Finally, the orders will be assumed to take between one and two minutes to arrive at the
kitting workstation as well as to travel to the exit from the kitting workstation.

Step 1: Create a new model that has one SOURCE (SrcOrdersArrive), one SINK (SnkOrdersLeave), and
one WORKSTATION object (WrkKitOperation) as seen in Figure 9.2.

· Connect the objects together using TIMEPATHS that take the appropriate Uniform(1,2) minutes as
the Travel Time property.
· The kits’ orders arrive with an Exponential(3) minute interarrival time.

The WorkStation Object: A Kitting Process 135

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 · The time needed to process each kit at the workstation follows a Triangular(0.1, 0.2, 0.3)
minutes distribution.
· The workstation’s tear downtime is one minute.
· The current setup time is a specific two minutes for the workstation.

Figure 9.2: Kitting Operation Model

Step 2: Insert a new MODELENTITY named EntOrders into the model.

Step 3: Add three additional symbols for the EntOrders entity by first selecting the EntOrders instance
and clicking the Add Additional Symbol button. Color the additional symbols “Yellow(1)”, “Red(2)”, and
“Blue(3)”122. Note that symbols are numbered starting from zero.

Step 4: All of the SIMIO objects have certain built-in characteristics. Properties for an object (e.g., initial
number of in system, maximum arrivals) are established at the time the object is created (usually at the
beginning of the simulation) and cannot change during the simulation. State variables are characteristics
that can change during the simulation. The SIMIO-defined Priority state variable123 is an example of a
characteristic that is often used for ranking. In SIMIO, you have the ability to add your own properties as
well as discrete state variables to objects.

Properties and States are added through the “Definitions” tab of an object. All objects in SIMIO can have
their own Properties and States. A state variable defined in the MODEL will only have one value and can be
thought of as having global scope. Defining a state variable on the MODELENTITY will allow each entity to
have its own state variable value. Select the MODELENTITY in the [Navigation] panel and select the
Definitions→States section. Each kit order will have a different quantity of kits requested and a state
variable is needed to track this information.

· Notice, two variables are already defined by SIMIO. Picture is defined by SIMIO as a DISCRETE
REAL STATE VARIABLE that represents which picture should be displayed to represent the entity.
Animation is also defined by SIMIO and is a STRING STATE VARIABLE which is used to determine the
type of moving animation the entity might do.
· You should add EStaOrderQuantity as new DISCRETE INTEGER STATE VARIABLE that will
represent the number in the order, as seen in Figure 9.3.

Figure 9.3: Add a State Variable to the MODELENTITY Object

122 Symbol numbers are in parentheses

123
There is an entity property called Initial Priority, which initializes the Priority state variable.
136 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 5: A table is needed to determine the percentage of time each type of order will occur as well as the
symbol picture number. First, select the MODEL in the [Navigation] panel to return back to the simulation
model. Let’s create the TableSymbol for the entity pictures as seen in Table 9.1 via the Data tab where
the Symbol column should be an Integer Standard Property while the Percent could be an Integer, Real
or Expression property (i.e. numeric).

Table 9.1: Order Type Symbol and Percentage

Step 6: Next, the order type (i.e., yellow, red, etc.), the order quantity for the arriving order, and the correct
picture will be specified using an Add-On process trigger. Select the SrcOrdersArrive and double click
the CreatedEntity Add-On process trigger property which will create the process named be
SrcOrdersArrive_CreatedEntity. In this process we can make assignments of state variables to the
incoming entities as shown in Figure 9.4 after setting the table row.124

Figure 9.4: Setting the Row of the Data Table and Assigning Picture.
Step 7: Use the Set Row process step to randomly assign a row of the TableSymbol TABLE to the entity
which will be used in the Assign step to specify the entity’s picture.

Figure 9.5: Steps for the OrdersArrive_CreatedEntity Process

Question 1: Why was the table set in the “Created” Add-On Process Trigger versus “Creating” as was
done in Chapter 5?
________________________________________________________________________________

Step 8: Next assign the correct picture and the order quantity. You can do this in two different assignment
steps or in one step using multiple rows by clicking the button on the “0 Rows”. You need to add each
assignment as seen below in Figure 9.6 through Figure 9.8.

· Use the “Repeating Property Editor” to make multiple assignments as shown in Figure 9.6.

124
Instead of using an add-on process trigger, the On Created Entity Table Reference Assignment could have been used to do the same
thing along with the Before Exiting State Assignments.
The WorkStation Object: A Kitting Process 137

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 9.6: Making Multiple Assignments
· Set the picture to represent the symbol associated with the specified order type as in Figure 9.7.125

Figure 9.7: Picture Assignment

· Assign the order quantity from a Binomial Distribution as in Figure 9.8.

Figure 9.8: Order Quantity Assignment

Question 2: Note a Binomial random variable is being used to model the number of orders. Why are we
adding one to the random variate?
________________________________________________________________________________

Step 9: Save your model with a new name and run the model for eight hours

Question 3: What is the average flow time for the orders?

________________________________________________________________________________

Question 4: How many orders were processed in the 8 hours?

________________________________________________________________________________

Question 5: What are we missing or what assumptions are we making in the current model?
________________________________________________________________________________

Part 9.2: Sequence-Dependent Setup Times

Whenever the kitting workstation must change from one color to another, there is a setup time that depends on
the color type that preceded the current order. The changeover time is shown in Figure 9.9.

Figure 9.9: Changeover Setup Times (From/To) in Minutes

125
The Picture state variable sets the symbol associated with the object in this case the ModelEntity.
138 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Let’s add the sequence-dependent setup times to our model. Under the WORKSTATION properties, you can
specify the setup times with the Specific constant setup (e.g., two minutes for the kitting operation). But
before we can use sequence-dependent setup times, we need to know how the changeover times can be
incorporated into the SIMIO model.

Step 1: First, a LIST of identifiers is needed that can be used to create the changeover table (matrix) where in
this case, a List of Strings. Under the “Definitions” tab, insert a LIST of Strings named LstColors. The
contents of the LstColors list would correspond to our symbol colors: “Green”, “Yellow”, “Red”, and
“Blue” as seen in Figure 9.10. It is important these are put in this order to correspond to the symbols (0, 1,
2, and 3).

Figure 9.10: List of Colors

Step 2: Now in the “Data” tab, add a CHANGEOVERS matrix (this is a “from-to” matrix) named Changeover
and use the LstColors list for the associated list. Change the unit of time to Minutes and enter the values
into the matrix from Figure 9.9.

Step 3: Next, specify the Setup time logic in the KitOperation WORKSTATION as seen Figure 9.11.

· The Setup Time Type property: Sequence Dependent

· Operation Attribute: TableSymbol.Symbol
· Changeover Matrix: Changeover

Figure 9.11: Specifying Sequence Dependent Changeovers in the Workstation

Step 4: We can now finish specifying the other elements of the Process Logic for the WORKSTATION as seen
in Figure 9.12.

· Operation Quantity property (i.e., the number of items in each order):

ModelEntity.EStaOrderQuantity
· Processing Batch Size: 1
(This is the number of items that can be processed at one time. So if the Operation Quantity is 10 and
the batch size is 2, it will take 5 batches to process the entire order with each batch requiring a
different random processing time.)
· Teardown Time: 1 minute

The WorkStation Object: A Kitting Process 139

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 9.12: Workstation Properties
Step 5: The workstation may be in several different resource states including starved (i.e., idle),
processing, performing a setup, or performing a teardown. To see all the nine possible workstation states,
click on the workstation and then click the “Active Symbol” to get the display in Figure 9.13.

Four more states

than the Server

Figure 9.13: Workstation States

Step 6: To see the percentage in these “resource” states evolve over time within the simulation; add a
“Status Pie” from the Animation tab. Specify the Data Type property as a ListState and the List State
should be the WrkKitOperation.ResourceState.126

Figure 9.14: Workstation Status Pie Chart

126Note, if you select the WrkKitOperation object first, the “Status Pie” will be associated with the WORKSTATION and you will only
need to specify ResourceState.
140 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 7: Save and run the simulation with animation for an hour. Then run it in fast-forward mode for two,
four, and eight hours. After eight hours, answer the following questions.

Question 6: What is average flow time for orders?

________________________________________________________________________________

Question 7: How many orders were processed in the 8 hours?

________________________________________________________________________________

Question 8: What are we still missing in the current model?

________________________________________________________________________________

Part 9.3: Sequence-Dependent Setup Times that are Random

The previous section utilized the CHANGEOVER matrix to handle sequence dependent setup times. However, the
matrix is limited to only constant/deterministic times and cannot handle the case when the setup times have
variability associated with them (i.e., random distributions). Since SIMIO’s CHANGEOVER matrix cannot utilize
distributions as the sequence dependent setup times, we will handle this manually using related data tables.
Similar to previous examples, a child table will be used to define the setup times for a particular part.

Step 1: Save the current model as a new name to be used again in later section.

Step 2: First we need to declare the relationship (i.e., primary key) for the parent table. From the
Data→Tables section, insert a new String Standard Property named Color. Click the Set Column as Key
button and fill in the field’s values as seen in Figure 9.15.127

Figure 9.15: Adding the Color Field as the Primary Key

Step 3: Insert a new DATA TABLE named TableChangeOvers as seen in Figure 9.16 which will be the
child table.

127
We could have used the Symbol field as the primary key but this will make the setups easier to visualize in the related table.
The WorkStation Object: A Kitting Process 141

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 9.16: Creating the Sequence Dependent Setup Times Table
Step 4: From the Add Column section, select the Foreign Key to insert the To Color column. Set the Table
Key property to TableSymbol.Color which will link the parent to the child changeover’s table.

Figure 9.17: Adding the “To Color” Child Foreign Key

Step 5: Insert a Standard Property named From Color and an Expression Standard Property named
ChangeOver. Make sure to set the Unit Type property to “Time” and the Default Units to “Minutes”.
Since this column is an Expression property column, we will be able to specify any SIMIO expression.

Step 6: Insert the values from Figure 9.16 which replicates the previous section’s setup times.

Step 7: Recall, when the EntOrders are created they are assigned a row to the TableSymbol (see Figure
9.4) which was used to determine the picture. Now, they will also be assigned the related portion of the
TableChangeOvers as seen in Figure 9.18 if the new order was “Red.”

GStaPreviousColor set to 2

Figure 9.18: Related Records in the ChangeOver Table Associated with Red Orders
Step 8: If the new order “Red” is next to be kitted and the previous order was “Yellow,” then the setup
time should be six which is associated with the second row of the related table as seen in Figure 9.18.
From the Definitions tab, insert a new MODEL Discrete Integer State variable named GStaPreviousColor
with a default value of “1” which will indicate the last color that was processed.
142 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 9.19: Insert a State Variable to Keep Track of the Last Order Processed
Step 9: Insert the FinishedSetup Add-on process trigger for the WrkKitOperation WORKSTATION which is
executed after the setup has completed. Insert an Assign step that will set the GStaPreviousColor to the
current symbol plus one which will be used for the next order. The symbols range from zero to three but
table rows start at one.

Figure 9.20: After Setup Completed Set the Previous Color

Step 10: Next, for the WORKSTATION, set the Setup Time Type back to “Specific” and specify the Setup Time
to be TableChangeOvers[GStaPreviousColor].ChangeOver. The GStaPreviousColor
indexes into the related table to specify the setup time.

Figure 9.21: Changing the Setup to Utilize the Sequence-Dependent ChangeOver Data Table
Step 11: Save and run the model comparing the results from the previous section.

Question 9: What is the time in system and the average setup time of the workstation?
________________________________________________________________________________

Step 12: To demonstrate the ability to have variable setup times, change all the setup times that are nine
minutes to Random.Pert(8,9,10) minutes. Save and run the model comparing the time in system
and the average setup time to the previous results.

Question 10: What is the time in system and the average setup time of the workstation now?
________________________________________________________________________________

Part 9.4: Using Materials in the Kitting Operation

The WORKSTATION has more capabilities than a general SERVER. For example, it can be used to keep track of
raw materials that have to be consumed, as specified in the “bill of materials” needed to create a part or
complete an assembly. If the raw material is not available, then parts (entities) cannot be processed and will
wait until the required materials become available.

Step 1: Return to the model from Part 2 of this chapter and save it as a new project.

Step 2: First, we need to declare the existence of raw materials that make up a kit as well as the Bill of
Materials that specifies the number of each raw material type needed to make a kit. Five new MATERIAL
The WorkStation Object: A Kitting Process 143

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 “Elements” need to be defined via Definitions→Elements→Workflow section as seen in Figure 9.22.128

· Create MatSubA, MatSubB, and MatConnectors as materials types, all with Initial Quantities of
zero, which will be changed shortly.
· Next create a MatBOMKits as MATERIAL element, but using the BillofMaterial Repeating Property
Editor as shown in Figure 9.22 to define what materials make up this product (i.e., one MatSubA and
MatSubB and two MatConnectors).

Figure 9.22: Five Materials and the Bill of Materials for the MatBOMKits
· Finally, create MatKits as the last “Material” element (with Initial Quantity of 0) which will
represent the finished product produced from the raw materials in the bill of materials.

Step 3: Since the initial inventory is a fixed characteristic of our model and only needs to be initialized at
the start of the simulation, here is an ideal place to use a SIMIO Model “property.” From the
Definitions→Properties” section, add new properties to objects. Create three new “Standard Real
Properties” named InitInvA, InitInvB, and InitInvC. Place all of these properties into a new InvModel
Category. The first time, the new category name (InvModel) needs to be typed into the window. For the
next two properties you can select the category from the dropdown.

Step 4: Return to the material “Elements” and now set the Initial Quantity property for each of the three
materials (i.e., MatSubA, MatSubB, and MatConnectors) as a reference property to the previously
defined properties. Remember this is done by right-clicking on the drop-down arrow and selecting the
referenced property ( ).

Step 5: As part of the analysis, create some statistics on the consumed materials as well as the produced
product. There are three different types of statistics that can be defined. Recall that a “Tally Statistic” is
an observation-based statistic (e.g., waiting time at a queue). A “State Statistic” is a time persistent
statistic on a state variable (e.g., number in queue). An “Output Statistic” can be used to collect any
expression value, once at the end of each replication. Add four OUTPUT STATISTICS elements with the
following names and expressions.

· OutStatSubAConsumed based on MatSubA.QuantityConsumed

· OutStatSubBConsumed based on MatSubB.QuantityConsumed
· OutStatConnectorsConsumed based on MatConnectors.QuantityConsumed
· OutStatKitsProduced based on MatKits.QuantityProduced

128Elements are objects that represent things in a process that change state over time (e.g., TALLYSTATISTIC, STATESTATISTIC,
TIMER, etc.)
144 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 · For each added “Output” statistic, under the Results Classification section set the Data Source to the
name of the material (e.g., “MatSubA”), the Category “Total Consumed” or “Total Produced” and
the Data Item “Number” as seen in the example of Figure 9.23 for MatSubA.

Figure 9.23: Setting of Results Classification

Step 6: Now return to themodel under the Facility tab and for the WORKSTATION KitOperation fill in the
Materials & Other Constraints properties (i.e., Material Consumption, Consumed Material Name,
Consumed Quantity, Material Production, Produced Material Name, and Produced Quantity) as seen in
Figure 9.24. This will force the WORKSTATION to consume materials based on the bill of materials as it
processes the kits.

Figure 9.24: Other Material Requirements for a WorkStation

Step 7: To access the new MODEL properties we created, click on the Model “Properties” (by right-clicking
the “Model” object in the [Navigation Panel]).129 In the Controls →InvModel section, set the values of
the InitInvA to 100, InitInvB to 100, and InitInvC to 200.

Figure 9.25: Setting Model Properties

Step 8: Save and run the new model for eight hours observing the statistics. While you may want to run it
for a while looking at the animation, you will need to fast-forward to the end to get the results.

Question 11: After looking at the Kits produced, do you have concerns about the model specifications?
________________________________________________________________________________

Question 12: How many kits were produced in the eight hours?
________________________________________________________________________________

Step 9: Change the initial inventory values to allow kits to be produced throughout the eight hours.

Question 13: How many orders were satisfied over the eight hours?
________________________________________________________________________________

129
Note you can change the name of the fixed model as well as set other model properties.
The WorkStation Object: A Kitting Process 145

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 14: How many kits were produced now?
________________________________________________________________________________

Question 15: How many of each raw material MatSubAs and MatSubBs were needed?
________________________________________________________________________________

Question 16: How many Connectors were needed?

________________________________________________________________________________

Question 17: What was the cycle time for the orders?
________________________________________________________________________________

Part 9.5: Raw Material Arrivals during the Simulation

Instead of material being available for the entire eight hour day at the beginning of the simulation, the raw
materials will be scheduled to arrive every two hours, beginning at time zero. When the stock arrives, the
inventory is returned to a target stock level – implementing an “order-up-to inventory policy”. How would
this be modeled in SIMIO?

Step 1: First, let’s collect some statistics on our “inventory” levels for the three raw materials over a given
day: MatSubA, MatSubB, and MatConnectors. Since inventory is a time-based statistic, add three
STATE STATISTICS in the Definitions→Elements section.
· StateStatSubAAvailable: MatSubA.QuantityAvailable
· StateStatSubBAvailable: MatSubB.QuantityAvailable
· StateStatConnectorsAvailable: MatConnectors.QuantityAvailable
· For each added “State” statistic, under the Results Classification section set the Data Source to the
name of the material (e.g., “MatSubA”), the Category “Inventory Level” and the Data Item
“Number”.
Step 2: Save and run the model for eight hours using initial inventory quantities of 1000, 1000, and 2000
for SubA, SubB, and Connectors (specified in the “Model Properties”).

Question 18: What was the average MatSubA available?

________________________________________________________________________________

Question 19: What was the average MatSubB available?

________________________________________________________________________________

Question 20: What was the average MatConnectors available?

________________________________________________________________________________

Question 21: How many kits were produced?

________________________________________________________________________________

Step 3: Now let’s implement a replenishment system in which a stocker comes by the kitting station every
two hours, beginning at time zero, and fills up the stock for MatSubA , MatSubB, and MatConnectors
to particular quantities.

Step 4: Add three order quantity properties for the model under the Definitions tab named
OrderQtySubA, OrderQtySubB, and OrderQtyConn. Each of these properties should be an Integer

146 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Data Format added to the InvModel category as shown in Figure 9.26. These properties will be used as
the order-up-to quantities that the stocker uses to replenish the raw material inventory.130

New Order
Quantity Properties

Figure 9.26: Order Quantity Properties

Step 5: Under the Definitions→Elements section, add a TIMER element which will fire an event based on a
time interval (a “timer event” will be an interruption in the simulation that allows a process to execute).
See Figure 9.27.

· Start the timer at time 0.0 (i.e., Time Offset property should be 0) meaning the first event fires at time
zero and it then should “fire” an event every two hours (i.e., Time Interval should be “2”).
· Name the timer TimerStockerArrives.

Figure 9.27: Timer for Stocker Arrivals

Step 6: Under the “Processes” tab, create a process named RespondtoTimerEvent that will respond to the
timer event when it expires as seen in Figure 9.28.

· Triggering Event property: TimerStockerArrives.Event

· Add an Assign step that makes the following assignments to the process which always make the
quantity available equal to the order quantity (i.e., order-up-to policy).131

MatSubA.QuantityAvailable = OrderQtySubA
MatSubB.QuantityAvailable = OrderQtySubB
MatConnectors.QuantityAvailable = OrderQtyConn

Figure 9.28: Process Used to Respond to the Timer Event

Step 7: Now in the “Model Properties” set the initial values of the raw materials InitInvA, InitInvB, and
InitInvC elements to zero, since the first TimerStockerArrives event occurs at time 0.0 and will
replenish the stock at the beginning of the simulation. Set the order quantities for OrderQtySubA,
OrderQtySubB, and OrderQtyConnectors to 250, 250, and 500 respectively.

130 Create one of the properties with the correct category and name. Then copy, paste it and change the name appropriately.
131 SIMIO provides a Produce step that essentially does the same thing and a Consume step which would perform a subtraction,

but the logic using an Assign step seems just straight forward. In this example, you would add three Produce steps that would
produce for example OrderQtySubA – MatSubA.QuantityAvailable.
The WorkStation Object: A Kitting Process 147

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 9.29: Setting the Model Properties for Order Quantities and Initial Inventory
Step 8: Save and run the model for eight hours and answer the following questions.

Question 22: What are the average MatSubA, MatSubB, and MatConnectors inventory levels?
________________________________________________________________________________

Question 23: How many kits were produced during the eight hours?
________________________________________________________________________________

Part 9.6: Implementing a Just-In-Time Approach

The previous example assumed a stocker would come by (exactly) every two hours and replenish the
inventory up to a certain level (i.e., an order-up-to policy). However, the company is moving toward a more
Just-In-Time (JIT) type of operation. They have a supplier that has located a warehouse very near the plant
and will respond to orders very quickly. The company has implemented a monitoring system such that when
the inventory drops below a certain threshold (i.e., reorder point), an order is placed to the supplier who can
supply the products within an hour.

Step 1: Save the current model to a new name. You need to delete the TimerStockerArrives TIMER
Element and the RespondToTimerEvent process or disable the TIMER by setting the Initially Enabled
property to False under the General properties of the TIMER.
Step 2: Add a new SINK named SnkInventory. Position it underneath the Kit Operation which will be
used to model the arrival of new raw materials (MatSubA, MatSubB, and MatConnectors).
Step 3: Add three reorder point properties for the model under the “Definitions” tab named
ReorderPTSubA, ReorderPTSubB, and ReorderPTConn. Each of these properties should be an
“Integer” Data Format added to the InvModel category. These properties will be used as the reorder
points for the JIT system to replenish the raw material inventory.
Step 4: Go to the “Definitions” tab and add three MONITOR elements underneath the “Element” section
named MonitorA, MonitorB, and MonitorC. A monitor element can be used to monitor the value of a
state variable when it changes value or when the state variable crosses some threshold (either from above
or below). Refer to Figure 9.30 for the MONITOR properties for MonitorA.

Figure 9.30: Monitor Properties

The following settings for MonitorA will monitor the SubA inventory level.
148 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 · Monitor Type should be CrossingStateChange.
· Crossing Direction should be Negative (i.e., 5 to 4 will fire an event if 4 was threshold).
· The Threshold Value should set to the new reference property we will call ReorderPTSubA.
· The MONITOR can cause a process to fire just as the TIMER element did but we will not use this feature
this time.

Step 5: Repeat the same process for raw material MatSubB and MatConnectors material using
ReorderPtSubB and ReorderPtConn reference properties respectively.

Step 6: Add three SOURCES (i.e., suppliers) named SrcSupplierA, SrcSupplierB, and SrcSupplierC
which will send the various raw materials to the kitting operation as seen in Figure 9.31.

Figure 9.31: Supplier Sources

· Assume that it takes between 50 and 65 minutes for the raw materials to be delivered between the
SOURCES and the SnkInventory SINK (this is modeled as Travel Time property of a TIMEPATH).
· Name each TIMEPATH as TPA, TPB, and TPConn respectively for the three TIMEPATHS.
· Set the Arrival Mode property of the suppliers to On Event and the Event Name to
MonitorA.Event, MonitorB.Event and MonitorC.Event respectively for each of the three
sources as seen for SrcSupplierA in Figure 9.32. Each source will create an entity when its monitor
event fires and sends the replenishment entity to the kitting operation.

Figure 9.32: Utilizing an On Event Arrival Process

Step 7: When the raw material reaches the end of the path, you need to increment the current quantity by
the appropriate order quantity using the ReachedEnd add-on process trigger for each of the three
TIMEPATHS. A Tokenized Process will be used instead of three separate processes. From the
Definitions→Token section, add a custom TOKEN named TknReplenish. For this TOKEN, add a “Material
Element Reference” state variables named TStaWhichMaterial, to reference the correct material and
“Integer” state variable TStaQuantity as seen in Figure 9.33

The WorkStation Object: A Kitting Process 149

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 9.33: Adding the Custom TOKEN to Pass in the Material and Quantity
Step 8: From the Processes tab, insert a new process named Replinish which uses the TknReplenish and
specifies two Input Arguments properties named MaterialToProduce and QuantityToProduce as seen in
Figure 9.34.

Figure 9.34: Specifying the Custom Token and Input Arguments

Step 9: Insert a Produce step (see Figure 9.35) which will be used to add the order quantity to the
particular material passed to the process. You will need to physically type the material name property
value TknReplenish.TStaWhichMaterial directly rather than selecting it from the dropdown list.

Figure 9.35: Incrementing the Quantity Available

Step 10: For reach of the TIMEPATHS (i.e., TPA, TPB, and TPC), specify the Replenish process as the
ReachedEnd Add-On Process Trigger property using the appropriate material and order quantity. Figure
9.36 shows the example for TPA using material MatSubA and quantity OrderQtySubA.

150 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 9.36: Using the Tokenized Process for Material A Production TPA Path
Step 11: In the Model properties, set the initial values of the raw materials MatSubA, MatSubB, and
MatConnectors elements back to 250, 250, and 500 since the replenishment is now being triggered. Set
the reorder points and order quantities to the values provided in Table 9.2.
Table 9.2: Initial Inventory, Reorder Point, and Order Quantities
Raw Material Reorder Point Order Quantity Initial Inventory
SubA 125 200 250
SubB 125 200 250
Connector 250 400 500

Step 12: Add six “Status Labels” from the Animation tab that will display the available quantities of each of
the three raw materials: MatSubA, MatSubB, and MatConnectors, perhaps as shown in Figure 9.37. The
first three status labels are just text with the other three labels having an expression as seen in in Figure
9.37 that will display the QuantityAvailable for each material.

Figure 9.37: Status Labels and the Expression for the Status Label
Step 13: Save and run the model for eight hours and answer the following questions.

Question 24: What are the average SubAAvailable, SubBAvailable, and Connectors available?
________________________________________________________________________________

Question 25: How many kits were produced during the eight hours?
________________________________________________________________________________

Part 9.7: Commentary

· This is one of the most interesting chapters in that it illustrates how to model what would be considered a
very complex operation composed not only of inventory concerns, but also supply chain issues. It’s a
powerful lesson and the topic of inventory and supply will be considered in additional chapters along with
some of the SIMIO features illustrated here.
· The use of MATERIALS and BILL OF MATERIALS is an extremely powerful modeling concept that
has numerous applications.

The WorkStation Object: A Kitting Process 151

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Chapter 10
Inventories, Supply Chains, and Optimization
Many industries/companies are facing many challenges in their effort to compete in the global marketplace.
Demand variations, long lead times, and raw material supply fluctuations can result in excessive finished
goods inventory and/or poor customer service levels. Many companies with longer lead times are forced into
a make-to-stock policy in order to respond to their customer demands quickly and maintain market share. Too
much inventory will create higher costs and may lead to product obsolescence or spoilage while to little
inventory may cause stock outs and lower customer service levels. Therefore, designing and optimizing a
company’s supply chain has become a priority as it is becoming a necessity for the survival of many
companies. The main objective of any supply chain is to supply the customer with a product when the
customer wants it and at a price that maximizes profit. In order to achieve both objectives, a balance has to be
struck between carrying sufficient inventories to meet demand, but not so much as to negatively impact
profitability.

Part 10.1: Building a Simple Supply Chain

Figure 10.1 shows a very simple three tier supply chain where customers arrive at a store (which could be a
collection of stores) to purchase a product. The stores place orders at the manufacturer DC (Distribution
Center) which will supply the product if it is in stock. The DC uses an inventory model to determine reorder
policies for the supplier, which is assumed to have a raw material source. It takes between three to six days to
truck the product from the supplier to the manufacturer.

Perfect
Raw Material Customers
Source
½ Day to Poisson(40)
Process Order Reorder
R d
Reorder Information
Information EDI Transaction
EDI Transaction 1 per Day

Product Product
Flow Flow
1 to 2 Days Once it Leaves DC
Supplier Owned By Store
Manufacturer
Manufacturer Store
Store
Triangular(10,15,20)
Triangular(10,15,20) Minutes
Minutes DC
DC
per
per Item
Item in
in the
the Order
Order Quantity
Quantity Pert(1,2,5) Minutes per
Item in the Order Quantity

Figure 10.1: Simple Three Tier Supply Chain

The goal is to track the DC’s average inventory and service level/fill rate over a year. Therefore, the store
customers will not be part of the current model since the manufacturer is interested in determining the reorder
point and reorder quantity. The store reorders daily from the DC and the DC utilizes a (s,S) periodic review
inventory model as seen in Figure 10.2. In this model, the inventory is reviewed every fixed time interval
called the period. An amount s is the reorder point, namely a quantity below which triggers a re-order. The
quantity S is determined as the quantity needed after inventory is reordered (an order-up-to amount).

152 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 (s, S) Inventory Policy
Inventory Level

Inventory Position = OnHand Inventory + OnOrder Inventory

S S (Order-up-to)

Inventory No Order

Lead Time (L) Amount Ordered

s s (Reorder Point)

Review Period

Time

Figure 10.2: (s, S) Inventory Policy

Step 1: Create a new model with two separate systems as seen in Figure 10.3.

· The first system will model the ordering portion of the supply chain (i.e., orders from the store placed
at the DC will need to be processed). Insert a SOURCE named SrcOrders, a SERVER named
SrvOrderProcessing and a SINK named SnkStore. A new MODELENTITY named EntOrders should
also be inserted making sure the SrcOrders creates these EntOrders. Use CONNECTORS to link the
objects together, modeling EDI transactions between the stores and the DC.
· The second system will model the flow of products from the back of the chain to the front of the
supply chain used to replenish the inventory in the manufacturer DC. Insert two WORKSTATIONS
named WrkSupplier and WrkDC and a SINK named SnkDCInventory along with another
MODELENTITY named EntReplenishments. Connect the WrkSupplier and WrkDC via a TIME PATH
that uniformly takes between one and two days. Use a CONNECTOR from the WrkDC to the
SnkDCInventory since the model is really only concerned with the DC.

Figure 10.3: SIMIO model of the Three Tier Supply Chain Reordering Process

Inventories, Supply Chains, and Optimization 153

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 2: Since we are not going to model every item in the inventory system as a separate entity, add a
DISCRETE INTEGER STATE variable named EStaOrderAmt to the MODELENTITY132 to represent the
quantity of product for each store order or replenishment reorder.

Step 3: For the Ordering process, set the following information.

· Change the SrcOrders to have one order arrive daily.

· The amount of each order that arrives daily from the store is Poisson distributed with a mean of 40.
As seen in Figure 10.4, use the State Assignments→Before Exiting property of the SOURCE to set the
ModelEntity.EStaOrderAmt to a Random.Poisson(40).

Figure 10.4: Specifying the Order Amount of each Order

Step 4: Next, under the Definitions→Properties of the model, insert four Expression Standard Properties
named InitialInventory, ReorderPoint, OrderUptoQty, and ReviewPeriod. These should all be placed
in the “Inventory” Category.133

Table 10.1: The Default Values for the Three Properties

Property Default Value Category Unit Type Description
The initial inventory at the start of the
InitialInventory 700 Inventory Unspecified
simulation.
ReorderPoint 300 Inventory Unspecified Determines the point to reorder.
The maximum inventory to order up to
OrderUptoQty 700 Inventory Unspecified
quantity.
ReviewPeriod 5 Inventory Time (Days) The periodic review period
Question 1: What are the advantages of setting up properties for this model?
________________________________________________________________________________

Question 2: How do you set the values for these properties of the Model?
________________________________________________________________________________

Step 5: In the Model, insert a DISCRETE INTEGER STATE variable named GStaInventory which will
represent the current inventory amount (“OnHand”) of the product at the DC and another one named
GStaOnOrder (i.e., work in process) which represents the total amount the DC has currently ordered
(“OnOrder”) from the supplier. The Initial State Value property should be 0 for both the GStaInventory
and GStaOnOrder variables respectively.

132 Select the MODELENTITY in the [Navigation] panel and go the Definitions→States section to add the state variable.
133
The first time, you will need to type the category name “Inventory” and then it can be selected from the dropdown box.
154 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 6: Next, we need to insert several ELEMENTS into the model as seen in Figure 10.5.
· Insert a TALLY STATISTIC named TallyStatSL to track the service level performance.
· Insert a STATESTATISTIC named StateStatInv to track the inventory level performance. Specify the
State Variable Name property to be the GStaInventory variable.
· Insert a TIMER named TimerReview to model the review period. The DC currently reviews their
inventory level every five days and should utilize the ReviewPeriod property (see Figure 10.5).

Figure 10.5: Setting up the Statistics and Periodic Review Timer

Step 7: Change the Processing Time property of SrvOrderProcessing such that orders take a ½ a day to
be processed before an attempt is made to fill the order with the available inventory.

Step 8: The supplier will take between 10 and 20 minutes with the most likely time of 15 minutes to
process each product. Therefore, the total processing time for one order would be the sum of a number of
triangular distributions (i.e., an order amount).134 The workstation handles the processing of the batch
correctly by sampling from the distribution the appropriate number of times. For the WrkSupplier,
specify the Operation Quantity property to be ModelEntity.EStaOrderAmt, the Processing Time to
be the individual process time (in minutes), and the Processing Batch Size property to “1” as seen in
Figure 10.6.135 Since the batch size is one, it will produce a set of sequential batches.

Figure 10.6: Specifying the Operation Quantity and Batch Size for Processing of an Order
Step 9: Once the product arrives at the DC, the entire lot must be individually packaged before it is
available to satisfy demands from the store. Therefore set the WrkDC’s Processing Time property to the
following expression Random.Pert(1,2,5)in minutes, the Processing Batch Size property to “1”
and the Operation Quantity property to be ModelEntity.EStaOrderAmt just like the
WrkSupplier.

134
One maybe inclined to use ModelEntity.OrderAmt*Random.Triangular(10,20,30) for the Processing Time
property. However, this will not correctly model the summed distribution as seen in the example in Appendix A.
135
If one does not specify the Processing Batch Size, it is assumed to be same as the Operation Quantity therefore processing the entire
quantity as only one batch rather a set of sequential batches.
Inventories, Supply Chains, and Optimization 155

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 10.7: Setting up the DC WORKSTATION Properties to Process the Batch
Step 10: Save and run the model for 26 weeks.

Question 3: How many store orders are produced during that time period and what is the average time in
the system?
________________________________________________________________________________

Part 10.2: Processing Orders in the Supply Chain System

Recall, the initial inventory level property(i.e., InitialInventory) was created to allow for changing the initial
condition at the model level.

Step 1: Since state variable default values cannot be reference properties, the state variable GStaInventory
needs to be initialized at the beginning of the simulation via the OnRunInitialized process of the model.
Insert this process by selecting it via the Processes→Select Process→OnRunInitialized drop down menu.
Insert an Assign step that assigns the GStaInventory state variable the InitialInventory as seen in
Figure 10.8.

0
Figure 10.8: Assigning the Initial Inventory Level
Step 2: When EntOrders arrive at the DC and have been processed, they need to be filled based on the
current inventory position.136 Therefore, insert a new “After Processing” add-on process trigger for the
SrvOrderProcessing SERVER to handle the behavior (i.e., update the current inventory position and the
current service level).

· Add a Tally step which will update the TALLY STATISTIC TallyStatSL with the value equal to
Math.Min(1,GStaInventory/ModelEntity.EStaOrderAmt).137
· Next, insert an Assign step that will update the current GStaInventory level by subtracting the
EStaOrderAmt of the particular order from it. However, if the GStaInventory level is less than the
order amount it would be negative and since we will not allow backorders set the level to the
following expression as seen in Figure 10.10.
· Math.Max(0,GStaInventory-ModelEntity.EStaOrderAmt)

136
Inventory position is defined as on-hand inventory plus on-order inventory (to account for orders that haven’t arrived).
137
Some may argue this service level calculation is really a fill rate, if the inventory is greater than the order amount then 100% of the
demand appears to be satisfied.
156 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 10.9: Process to Fill Demands from the Stores

Figure 10.10: Properties of the Steps

Step 3: To track the information while the simulation is running insert eight status labels from the
“Animation” tab, as seen in Figure 10.11, that keeps track of the current level, the average inventory and
service level. The first column of labels is the description while the second column uses the expressions
GStaInventory, GStaOnOrder, StateStatInv.Average, and TallyStatSL.Average.

Figure 10.11: Status Labels to Show Current Inventory Level and Average Inventory and Fill Rate
Step 4: From the “Animation” tab, insert a new STATUS PLOT that will track the current onhand inventory
level, the replenishment on order, and the inventory position which represents the onhand plus the on
order. Using the Additional Expression property, add three items to the plot as seen in Figure 10.12 and
make the Time Range equal to “1” week.

Figure 10.12: Setting up the Status Plot to Track Inventory and On Order Amounts
Step 5: Save and run the model (you may need to adjust the speed factor to 5000 or fast forward).

Question 4: What is the current inventory level?

________________________________________________________________________________

Question 5: What is the average inventory level and service level?

________________________________________________________________________________

Inventories, Supply Chains, and Optimization 157

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Part 10.3: Creating the Replenishment Part of the Supply Chain System

The previous section handled demand orders from the stores but once the inventory was depleted, there was no
replenishment of the inventory. Hence, the service level was very poor. The timer was setup to fire every five
days, at which time if the current inventory is below the reorder point, a replenishment order should be sent to
the supplier based on the order-up-to quantity, the current inventory level and the current number of
outstanding orders (On Order). Recall that the “inventory position” is defined as the current inventory on hand
plus the inventory represented in outstanding orders.

Step 1: Create some statistics on the replenishment procedure:

· Insert a STATE STATISTIC, called StateStatOnOrder, to keep track of the amount of inventory on
order over time. This statistic should reference the GStaOnOrder state variable.
· Insert a TALLY STATISTIC, called TallyStatAmtOrdered, to record the amount that is ordered from
the supplier.

Step 2: Setup the periodic review by creating a process to respond to the timer TimerReview. From the
“Processes” tab, create a new process named InventoryReview. Set the Triggering Event property to
respond to the TimerReview.Event which causes the process to be executed every time the timer
event fires as seen in Figure 10.13.

Figure 10.13: Process to Reorder

· Insert a Decide step that uses a “ConditionBased” check to see if the current inventory is less than
the reorder point (i.e., GStaInventory < ReorderPoint).
· If a new reorder appears needed (i.e., “True” branch), use the Create138 step to create a new
EntReplenishments entity as seen in Figure 10.14.

Figure 10.14: Creating a New ReOrder to Send the Supplier

· For the newly created EntReplenishments, use an Assign step to assign the order amount and then
increase the StaOnOrder value based on the order amount.

138
The Create step can create new objects or copies of the associated or parent objects. Each of the created objects will be executed
by their own token. Note that created tokens are updated in SIMIO before the original.
158 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Table 10.2: Determining the Order Amount and Updating the WIP
State Variable New Value
ModelEntity.EStaOrderAmt Math.Max(0,OrderUpToQty –GStaInventory +GStaOnOrder)
GStaOnOrder GStaOnOrder + ModelEntity.EStaOrderAmt
· In the second Decide, determine if the EStaOrderAmt is zero. Namely, no replenishment order is
needed (an assumption here is that there are never any zero quantity replenishment orders)
· If no replenishment order is needed, then the replenishment order entity is destroyed using the
Destroy step.
· Otherwise, the amount of the replenishment order (EStaOrderAmt) is tallied via the
TallyStatAmtOrdered.
· Once the EntReplenishment has been tallied, it needs to be sent to the supplier by using a
Transfer step to move it from “FreeSpace” to the Input node of the WrkSupplier.139

Figure 10.15: Sending the New Reorder to the Supplier

Step 3: Now, when the products arrive back at the DC and have been packaged, the GStaInventory
variable needs to be increased by the order amount while the GStaOnOrder variable is reduced by the
same value. Insert the “FinishedGoodOperation” add-on process trigger for the WrkDC as seen in Figure
10.16. Insert an Assign step that updates the values in Table 10.3.

Figure 10.16: Updating the Inventory and WIP Values

Table 10.3: Updating Inventory and DC once Product Arrives

State Variable New Value
GStaInventory GStaInventory + ModelEntity.EStaOrderAmt
GStaOnOrder GStaOnOrder - ModelEntity.EStaOrderAmt

Step 4: Save and run the model (you may need to fast forward the simulation) for 52 weeks.

Question 6: What is the average inventory level, the inventory on order and the service level?
________________________________________________________________________________

Question 7: What is the average amount ordered?

________________________________________________________________________________

Question 8: Is this an adequate system?

________________________________________________________________________________

139
When entities are created, they are placed in “FreeSpace” and, typically, will need to be “transferred” from there to one of the model
nodes or stations.
Inventories, Supply Chains, and Optimization 159

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 5: You may notice that the status labels display as many significant digits that are available for both
the average inventory and the service level/fill rate. SIMIO exposes many of the .Net string manipulation
functions. For example, String.Format function can be used to format the output.140 Replace the two
status labels for the average and service level with the ones in Table 10.4.

Table 10.4: Formatting the Output of the Status Labels

Status Label Expression
Average Inventory string.format("{0:0.##}",StateStatInv.Average)
Average OnOrder string.format("{0:0.##}",StateStatOnOrder.Average)
Service Level Status string.format("{0:0.00%}",TallyStatSL.Average )
Step 6: Save and run the model (you may need to adjust the speed factor to 100 or fast forward the
simulation).

Question 9: Was the average inventory level, on order, and service level now formatted correctly?
________________________________________________________________________________

Part 10.4: Using an Experiment to Determine the Best Values

Simulation is often used to improve a system. Can we find a better order-up-to quantity and re-order point that
improves the overall performance of the system? Can SIMIO provide help with this kind of question? Or can
SIMIO help in searching for an improved system? The answer is yes to all these questions.

Step 1: From the Project Home→Create section, insert a new “Experiment” named FirstExperiment
which will create a new experiment window and automatically create the first scenario which is the base
model as seen in Figure 10.17. SIMIO automatically added the four properties as control variables.141

Step 2: Insert eight more scenarios by clicking the little box in the last row. This will copy all of the
parameters from the base case. Then change reorder points and order up to quantities to do a full factorial
experiment of the three reorder points (200, 300, and 400) and order up to quantity (600, 700, 800) as seen
in Figure 10.17. Change the names of the scenarios so they can be interpreted.

Click here create a

copy of the base
case scenario

Figure 10.17: First Experiment Trying Different Reorder Points and Order Up to Quantities

140 The syntax for the format function is String.Format(string, arg1, arg2, …) where the format “string” can contain
place holders that are enclosed by {} which start with the 0 th argument (i.e., {0}). For example, String.Format(“SL = {0} and
the Avg = {1}”, 0.45, 89.4) has two arguments {0} and {1} and replaces those with the numbers 0.45 and 89.4 respectively
to produce a ‘SL=0.45 and the Avg = 89.4’. Along with the arguments, one can specify formatting information especially as applied to
numbers by using a colon on the argument then specifying the number format. For example, {0:#.##} will display only two significant
digits while {0:#.00%} will display a percentage and force there to always be two significant digits.
141
Other control variables can be defined.
160 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 3: Using the Design→Experiment→Add Response button, insert three responses which will be used
(see Figure 10.17) to select the best scenario with the following parameters specified in Table 10.5. The
first two responses seem very logical. In this model these responses are typically conflicting (i.e., as the
inventory level is minimized the service level goes down). Therefore, the third one converts the multi-
objective problem into a single response. As the service level goes up and the inventory goes down, this
new response goes down. The lower and upper bounds on the responses will be used to flag responses
outside a stipulated range.142 In this case we would like to have a service level greater than 90%.

Table 10.5: Experimental Responses

Name Expression Objective Lower Bound
AvgInventory StateStatInv.Average Minimize
Service_Level TallyStatSl.Average Maximize 0.90
Inv_SL StateStatInv.Average/TallyStatSL.Average Minimize

Step 4: Figure 10.18 shows the basic experiment parameters where you can define the warm up period, the
default number of replications143 that all scenarios will run. For the SMORE plots and optimization, the
confidence interval and upper and lower percentiles displayed. Also, you need to define the primary
response that will be used by the ranking and selection and Optquest™ add-ins (see next sections).

Figure 10.18: Setting up the Experiment Parameters

Step 5: Save the model and run the experiment.

Question 10: Based on the responses, which reorder point and order up to quantity seems to be the best?
________________________________________________________________________________

Question 11: How did you make this determination?

________________________________________________________________________________

Part 10.5: Using SMORE Plots to Determine the Best Values

In the previous section, we just used the average of the ten replications to assist you in choosing the best
selection. The difficulty is that the variability of the system is not accounted for when choosing the best
scenario. “SIMIO Measure of Risk & Error” (SMORE) plots offer the ability to see the variability, along with
confidence intervals. They show the major percentiles (median or 50th percentile, lower chosen percentile and
upper chosen percentile) and mean, and confidence intervals for the mean and the percentiles.

Step 1: Select the “Response Results” tab from the previously run model to see the SMORE plots. The
primary response (Inv_SL) will be displayed by default but the other response can be selected from the
dropdown as seen in Figure 10.19. Quickly we can see that from the “AvgInventory” response

142
The values in the response that don’t fall within the bounds are shown in a red gradient color.
143
The default is ten replications, although more replications will increase your confidence in the results.
Inventories, Supply Chains, and Optimization 161

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Scenario200-600 seems to be the best and its variation is clear of the other plots. Scenio300-600 appears
to be the second best; however its variation overlaps with Scenario200-700. However when looking at the
service level with the limits turned on, only Scenario300-600, Scenario400-600, Scenario400-700 and
Scenario400-800 meet the 90% service level limit with the Scenario400-700 seeming statistically the best.
Considering the two plots together, Scenerio300-600 appears the better choice.

Figure 10.19: SMORE Plots of the Average Inventory and Average Service Level
Step 2: Looking at the Primary Response (“Inv_SL”) SMORE Plot in Figure 10.20, Scenario300-600
appears to be statistical the best, based on the combined response. Its SMORE plot does, however,
overlap with Scenario400-600 minimally but not the confidence intervals.144

Figure 10.20: SMORE Plots of the Combined Inventory/Service Level Response

Step 3: By clicking the “Subset Selection” option in the “Analysis” section of the ribbon in the “Design”
tab, you can see the choices made by a series of algorithms within SIMIO. Scenarios for each response
are divided in to the “best possible group” and the “rejects group”, as shown in Figure 10.21.

144
Refer back to Chapter 8 for more explanation of SMORE plots and their meanings.
162 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 10.21: Using Subset Selection
While the best possible group may consist of scenarios that cannot be proven statistically different from
each other, it can be shown that the best possible group scenarios will be statistically better than the rejects
group (shown in muted color).

Question 12: Which scenario(s) is(are) statistically selected as the best for the three responses?
________________________________________________________________________________

Part 10.6: Using Ranking and Selection to Determine the Best Scenario

The analysis is quite simple (focused on only two variables within the model) and ten replications seemed to
be enough to choose the best. However, using the SMORE plots still requires a judgment or visual
identification. Ranking and selection methods allow you to determine statistically which scenario is the best.
One of the advantages of SIMIO is the built in state-of-the-art ranking and selection method based on research
by Kim and Nelson (called “KN”).145

Step 1: Save the current model and copy the FirstExperiment by right clicking it in the [Navigation] panel
and choosing the Duplicate Experiment menu item. Name the new experiment KNExperiment.

Step 2: Next, choose the “Select Best Scenario using KN” from the Design→Add-Ins section seen in Figure
10.22.

Figure 10.22: Selecting the KN Algorithm Add-in

Step 3: This will add the KN algorithm to the experiment (see Figure 10.23) as well as give you additional
properties.146 Set the Indifference Zone property to five, which stipulates that one solution, can be
determined better than another only if the two solutions differ by more than five. Since the ranking and
selection scheme will change the number of replications during its execution, the Replication Limit
property specifies the maximum number that can be run.

145
S. Kim and B. L. Nelson, "A Fully Sequential Procedure for Indifference-Zone Selection in Simulation," ACM Transactions on
Modeling and Computer Simulation 11 (2001), 251-273.
146
To access the EXPERIMENT properties, right click the experiment in the [Navigation] panel.
Inventories, Supply Chains, and Optimization 163

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 10.23: Setting up the KN Parameters
Step 4: Set the number of “Required” replications (under the Replications category) each to one and make
sure that each scenario is “checked” (in left column) and “Reset” the experiment.

Step 5: Run the experiment and notice that the number of replications required for each scenario increases
to ten. The KN algorithm requires at least ten replications in order to reliably detect a difference among
scenarios.

Question 13: Based on the differences among the scenarios Inv_SL, do you think a large number of
replications are needed to determine differences greater than five?
________________________________________________________________________________

Question 14: What does running more replications do for the analysis?
________________________________________________________________________________

Step 6: The KN algorithm has selected Scenario300-600 as the best (it is the only one checked), as seen in
Figure 10.24. In more complex models with scenarios that are more similar, more replications than ten
will be needed.

Figure 10.24: Results of running the KN

Step 7: Verify that Scenario300-600 is selected and examine its SMORE plot in the “Response Chart”.

Question 15: What is the 95% confidence interval on the mean?

________________________________________________________________________________

Step 8: Reset the experiment and make sure to select all the scenarios again. Change the indifference zone
to one, and rerun the experiment to see the impact.

Question 16: Did KN need more than ten replications to distinguish the best scenario and which scenario
did it select this time.
________________________________________________________________________________

164 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Part 10.7: Using OptQuest™ to Optimize the Parameters

The KN algorithm used in the previous section determined the optimal scenario from a fixed set of scenarios
defined by the user. SIMIO has another add-in optimizer OptQuest™ that can perform traditional optimization
by trying other values for the decision variables not specified. It will also allow constraints on the decision
variables and output responses.147

Step 1: Save the current model and duplicate the FirstExperiment and then rename the new experiment
OptQuestExperiment.

Step 2: Use the “Clear” button to clear out the Select Best Scenario add-in and then select the “OptQuest
for SIMIO” add-in if you duplicated the KNExperiment instead.

Step 3: Since the optimizer will try different values for our control (decision) variables, we need to define
their valid ranges. Select appropriate ranges and increment values to speed up the convergence and
potentially the solution quality. Select each of the control variables and set their values according to Figure
10.25. As you can see, the InitialInventory is not included in the optimization. An increment value of 25
has been chosen for the other two variables to allow the algorithm to only select values in increments of
25.148

Figure 10.25: Setting the Lower and Upper Bounds on the Decision Variables

Step 4: Next, choose the “OptQuest™ for SIMIO” from the Design→Add-Ins section to insert the
optimization algorithm. It will add a setup section in the EXPERIMENT properties as seen in Figure 10.26.
The algorithm does use a confidence level to distinguish among solutions. The relative error percentage of
the confidence level is expressed as a percent of the mean. Notice, that a range of replications is specified
and a maximum number of scenarios is specified. These will limit the time OptQuest will spend
computationally in the simulations and search for a better solution.

147
OptQuest is an additional add-in that has to be purchased.
148
Finer resolutions (i.e., smaller values) may lead to better solutions. However, this can affect the solution quality of the optimization
algorithm. Having a course resolution quickly allows the algorithm to find a good region. A second optimization with a finer resolution
could be run around the point found by the first optimization problem.
Inventories, Supply Chains, and Optimization 165

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 10.26: OptQuest Parameters
Step 5: Save and run the optimization algorithm. Notice how it tries several different variable values of the
controls. It will create a maximum of 100 scenarios or decision points as specified the Max Scenarios
property.

Step 6: Notice only scenarios which are feasible with respect to constraints on the responses (i.e., service
level greater than 90%) are checked. Click on the little funnel ( ) on the upper right corner of the
Service_Level response and choose the [Custom] filter option. Setup the filter to only allow feasible
scenarios to be visible as seen in Figure 10.27. Then sort the scenarios by the Inv_SL column in
ascending order to see those with minimum Inv_SL ratios by right clicking on the column heading.

Figure 10.27: Filtering only Solutions which meet the Threshold of 90% Service Level
Question 17: What appears to be the best reorder point and order up to quantity values considering not only
the Inv_SL ratio but the service level?
________________________________________________________________________________

Step 7: Since the OptQuest™ only used five replications; it is possible that the variability may be a
problem. Keep the top ten (or some other number) scenarios and delete the rest. Since the optimization
parameters (i.e., bounds on the control variables will be lost), first duplicate the experiment and name it
OptQuestExperiment2 by right clicking on the name in the [Navigation] panel.

Step 8: Select the OptQuestExperiment from the [Navigation] panel and clear the OptQuest add-in and
add in the Select Best Scenario and make sure the Primary Response is Inv_SL. Use an Indifference Zone
property of “1” and then rerun the KN algorithm.

166 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 18: What did you get for the best Reorder Point and Order Up to Quantity?
________________________________________________________________________________

Question 19: What alternative do you choose if you want the service level to be at least 90%?
________________________________________________________________________________

Step 9: To see that one is better than the others, examine the Response Chart for the one scenario KN
selected, as seen Figure 10.28. Note you must “check” the alternative scenarios to see them in the
Response Chart. In our example, Scenario 052 (i.e., 300 reorder point and 550 (order-up-to quantity) was
deemed to be the best with one as the indifference zone.

Indifference Zone of “1”

Figure 10.28: Viewing the Best Scenaro

Part 10.8: Multi-objective and Additional Constraints using OptQuest™

While the KN selection algorithm can statistically determine which scenario is the best, it cannot consider
constraints on the responses. Constraints on the control variables should be handled since the scenarios
selected should be feasible. Unfortunately, many problems often have multiple conflicting objectives and it is
rare that a single point optimizes all the objectives. Figure 10.29 depicts the Efficiency Frontier of a problem
which is optimizing both service level and inventory level at the same time. The frontier represents all non-
dominated points that a decision maker could choose. From the figure, the inventory level could be reduced in
half if they would be willing to go from 98% service level to around 90% service level.

100
90
Service Level (%)

50

Search Direction

10

0
5000 10000
Inventory Level

Figure 10.29: Efficiency Frontier

Inventories, Supply Chains, and Optimization 167

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Finding the frontier is very difficult. To help alleviate the problem, many people try to convert the multi-
objective problem into a single objective (g(x)) problem. The most common way and easiest way is to
aggregate the objectives into a single objective by summing up the weighted objectives (i.e.,
). The decision maker can determine the importance of each objective by choosing the
weights accordingly. SIMIO has the ability to automatically create the combined function by specifying the
appropriate weights of the responses and the choosing the Multi-Objective Weighted value as the Objective
Type property of the EXPERIMENT as seen in Figure 10.30.

Figure 10.30: Using the Multi-Objective Aggregate Optimization Method

In this example, the decision maker is putting twice as much importance on service level than inventory. The
difficulty with this multi-objective approach is scaling the weights to find solutions. In this example, inventory
is in the hundreds while service level is less than one. In many cases, the optimization algorithm will only
optimize the dominate objective. While this may be a convenient method, one can accomplish the same thing
by creating an aggregate response and specifying the expression to be StateStatInv.Average + 2 *
TallyStatSL.Average. In order to avoid the scaling problem, we created an aggregate response that
was minimized by taking the inventory level divided by the service level to try to force the algorithm to
minimize the inventory while maximizing the service level.

The other way is to optimize one primary objective (e.g., average inventory) while setting thresholds (i.e.,
lower and/or upper bounds) on the other objectives (e.g., lower bound of 0.90 on service level) which would
create one point on the frontier which we did in the previous section. There are two categories of constraints in
the OptQuest for SIMIO: constraints on the controls (inputs) and constraints on the responses (outputs).

Suppose we consider a slightly different optimization problem, namely minimizing the average inventory level
while keeping the service level above 95%. We can write this more formally as the following equations. Of
the three constraints, the last two are simply bounds on the controls. The first constraint restricts a response of
the simulation (i.e., Service Level).

Minimize: AvgInventory
Subject to: Service Level >= .95
200 <= ReorderPoint <= 500
500 <= Order UpToQty <= 1200

Step 1: Select the OptQuestExperiment2 and change the primary objective to be the AvgInventory
response under the experiment properties.149

149
Right click the experiment name in the [Navigation] window to access the properties of the experiment.
168 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 10.31: Change the Primary Response of the Experiment Properties
Step 2: Save and run the optimization algorithm and compare the results to the previous experiment.

Step 3: Sort the scenarios in the average inventory column in ascending order to see those with minimum
average inventory as well as filter the service levels to only display feasible solutions.

Question 20: What appears to be the best reorder point and order up to quantity values considering not only
the average inventory but the service level?
________________________________________________________________________________

Step 4: All of the previous methods performed the decision before the search. OptQuest can discover the
efficiency frontier as well. Duplicate the OptQuestExperiment2 and name it OptQuestExperiment3.

Step 5: Remove the Inv_SL response by selecting it and clicking the Remove Response button. SIMIO will
generate the frontier based on all responses present.150

Step 6: Change the Objective Type property under the OptQuestExperiment3 properties to “Pattern
Frontier” and the Max Scenarios to “200” as seen in Figure 10.32.

Figure 10.32: Changing the Objective Type to Pattern Frontier

Step 7: Select the Service_Level response and change the lower bound to 0.5 so we will generate points
above 50% service level.

Figure 10.33: Change the Lower Bound on the Service Level

150
The frontier for two responses is a line and difficult generate while three would need a plane needing many scenarios.
Inventories, Supply Chains, and Optimization 169

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 8: Save and run the model. After the model runs, sort the checked column in descending order to
show all the points that were created on the frontier as seen in Figure 10.34 while Figure 10.35 shows the
frontier graphed by copying the data to Excel. The data in Excel has had its duplicates removed and been
sorted in descending order.

Figure 10.34: Part of the Points Generated on the Pattern Frontier by SIMIO

Figure 10.35: Pattern Frontier Graphed in Excel

Step 9: Note that more complicated constraints can be composed on the inputs (controls). In fact, these
constraints are more common. A nonsensical example of a constraint on an input in this model could be
(ReorderPoint + 3*OrderUpToQty >=1000) which is inserted by selecting the Add Constraint
button and entering the equation as seen in Figure 10.36.

Figure 10.36: Example of Constraint on Decision Variables

Part 10.9: Commentary

· The use of the subset selection facilitates finding a good collection of solutions. For a large number of
control variables and scenarios, the KN ranking and selection scheme is quite good in determining
statistically the best scenario.
· OptQuest is a very good optimizer given enough time and replications. However the SMORES,
subset selection, and KN add greatly to determining what “solution” you think is best.
· With the Cost Financials added to SIMIO, one could have optimized cost as well.
170 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Chapter 11
Simulation Output Analysis
You have carefully created a model and paid close attention to the present operational system. Your model
produces appropriate performance measures. You may have done some considerable re-modeling as you
learned more about the system being modeled and its important aspects. You have been concerned about
modeling errors and checked your model often and verified it is working the way you want. Furthermore, you
have attended to the question of model validity as you have been checking the simulated performance
measures against actual results. Through all your efforts, you have been thinking about the ways to improve
the system performance and have an accumulated set of alternative scenarios you want to try. You are now
satisfied that you can begin to examine these alternative scenarios and begin the process of developing
recommendations.

The simulation model produces outputs but we know they are sets of random variables. But how do we best
understand these results. We can generate averages of observed outputs, but these averages exhibit variation.
Can we simply make multiple replications of our model, create confidence intervals, and draw conclusions?

Part 11.1: What can go wrong?

There are two overriding statistical issues that should concern us. The first is bias in our estimates and biased
estimates are potentially wrong estimates. The second is the “independence and identicalness” assumptions
which plague us in calculating measures of variability and ultimately comparing systems.

Bias

Perhaps the biggest concern is the potential “bias” in our computed averages. Bias means that we computed a
value that is systematically different than the performance measure of interest. In other words, we may be
computing a value that simply isn’t the value of interest. The bias can occur in our method of calculation, but
more likely it is in the values being used. For instance, consider the computation of the average waiting time
in the queue of the ice cream store. Now, we know that the simulation starts “empty and idle”, so that the
early waiting times are going to be considerable smaller than those during the day. In fact, expected waiting
time changes throughout the day. Since the arrival process uses a time-varying arrival, the waiting times grow
and shrink with strong influence from the arrival rate. Other statistics like resource utilization will vary as
well.

Bias may or may not be a problem. If we need to know the waiting time, say during the busiest time in the ice
cream store, computing the waiting time during the entire day will clearly underestimate that waiting time. In
such a case we need to collect our own statistics during the periods of interest. On the other hand, an overall
average waiting time computed over the whole day, can be useful. While it may underestimate our intuition
about the congestion, it is a systematic and consistent computation that can be useful in comparing alternative
scenarios. Because the waiting time will be computed in the same manner in all scenarios, the observed
improvement or deterioration will be “relatively” correct. In other words, our absolute estimate of waiting
times may be an underestimate, but that same degree of underestimation will be present in all scenarios
explored.

This idea that relative differences are the important perspective that allows us use many simulation models that
underestimate many performance measures simply because all the “small” factors that impact the performance
measure is being ignored. For instance, our simulation model may only account for 50% of a resource’s duties
when we know their workload is closer to 80%. But because all our scenarios ignore those same small factors
(e.g., meetings, breaks, interruptions), if we find a scenario that enhances the resource to 60% utilization, then
we believe that the base utilization of the resource will be increased by 10% by employing this scenario. So

Simulation Output Analysis 171

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 the recommendation here is to use a relative difference perspective rather than an absolute perspective when
possible.

Independent and Identical

In order to measure variability associated with any simulation output, one needs data to be independent and
identically distributed (IID). The reason is that the computation of the confidence interval for an expected
value is based on the central limit theorem, which allows us to compute the standard deviation of the mean
from the standard deviation of the observation. And unfortunately, variability is present in almost all our
simulation output, so we had to “construct” IID observations. We recognized early in our study of simulation,
that the observations within a single replication do not satisfy the IID requirement. So we rejected the
observed waiting times within a simulation replication as unsuitable for estimating the variance of an expected
value. Instead we compute the replication average as an observation and that works for most of the cases of
interest.

There are a couple of situations where getting the IID observations maybe a cause for concern. One case
occurs when a simulation replication takes a very long time. It becomes very costly in terms of computer
time)to do the appropriate replications needed. A second case is when the simulation starts from, for example,
empty and idle and there is a lengthy period when the performance measures are poorly estimated due to the
startup of the simulation. In both cases, using replications (i.e., one observation per replication) becomes
problematic since we have limited opportunity to do many replications.

Part 11.2: Types of Simulation Analyses

There are two general types of simulation analyses: terminating and steady-state. A terminating simulation
typically models a terminating system that has a natural opening and closing. The ice cream store is a good
example because it opens empty and idle in the morning and then closes in the afternoon, sometimes after the
last customer is served. Terminating simulations are characterized by their opening and closing. Thus any
performance measures computed are implicitly dependent on how the simulation is started and stopped. One
replication is generally the time between the opening and closing. A terminating simulation is sometimes
called a “transient” simulation because its state is constantly changing throughout the replication. Although
bias in the estimates of the averages may occur because of the changing state, the estimation of variance is
facilitated by the use of replications.

Question 1: What is an example of a simulation that should be examined as a terminating simulation?

________________________________________________________________________________

A steady-state simulation has no “natural” opening or closing. In a steady-state simulation, we want our
performance measures to be independent of how we started or stopped the simulation. In other words, we
want a kind of long-term average for the expectations. However, while there is considerable appeal for
performance measures that independent of starting and stopping, just how to obtain such a simulation is not
clear. There are two fundamental problems. The first is how to eliminate the effects of startup or warmup.
The influence of the way the simulation starts can extend far into the simulation. The second problem is how
long a replication should last, since there is no natural stopping state. Neither of these problems can be easily
answered, so the tendency is to resort to “rules of thumb” that give some general guidance, but for which there
is limited theory.

Question 2: What is an example of a simulation that should be examined as a steady-state simulation?

________________________________________________________________________________

Part 11.3: Output Analysis

To provide a context for an output analysis, re-consider the work cell problem of Chapter 5, section 6. Recall
that this problem is a manufacturing cell consisting of three workstations (i.e., A, B, C) through which four
172 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 part types are processed. The parts have their own sequences and processing times depended on the part type
and the station visited. We defined two statistics of interest: a “state statistic” on the number of Part type 3’s
in the system during the simulation and a “tally statistic” on the time Part type 3’s are in the system.

Very little analysis of the problem was undertaken. We ran the simulation (interactively) for 40 hours and
examined the output. The average number of Part type 3’s in the system was 0.40 parts and the average time
for Part type 3’s was in the system was 0.26 hours. At that time, our focus was on the construction of the
model not necessarily the interpretation of the output which will explore now.

Since we have more experience with simulation output, we may be considering an SIMIO experiment and
making multiple replications. By making multiple replications, we obtain an estimate of the variability
associated with the observed averages and could compute a confidence interval. But before doing that
analysis, let’s think about our simulation intent. How do we want to use the output from the simulation? Do
we want to say something about the “short-term” or the “long-term” outcomes? We probably chose a 40 hour
simulation length because it represents a “weeks’ worth” of work. However, we also used the default “empty
and idle” state for the start of the week. Is that representative of a real week? If the answer is yes, then we can
proceed to making multiple replications and the creating confidence intervals for our output measures.

Question 3: In your judgment, how likely is it that you are interested in one week statistics being empty
and idle at the beginning of the week?
________________________________________________________________________________

By doing replications we should obtain approximate IID observations, but a more fundamental concern is with
the bias. We have two sources of bias. The first is the “initialization” or “warmup”. Starting empty and idle
certainly affects our statistics. The second is whether the 40 hours is sufficient time to observe the system. To
limit our let’s focus primarily on the two user-defined statistics: the number of Part type 3’s in the system
during the simulation and the time Part type 3’s are in the system.

Initialization Bias

When we describe the initialization bias, we often use the term “warmup” implying that the simulation output
goes through a phase change like a car that needs to warm-up in cold weather before turning on the heater.
We think of a “transient phase” followed by a transition to the “steady-state phase.” And usually, it’s the
steady-state phase that interests us since it represents the “long-term” perspective.

Three “obvious” ways to deal with the initialization bias are: (1) ignore it, trusting that you have enough
“good” statistical contributions from the steady-state phase that the “poor” statistics at the beginning (during
the transient period) are completely diluted, (2) “load” your system with entities at the beginning of the
simulation, so that the transition to steady-state behavior occurs early, and (3) after a while (i.e., warm up
period), clear (truncate) the statistics in an attempt to clean out the transient statistics so they can no longer
influence the statistics collection.

The problem with ignoring the presence of the transient statistics is that their influence on the final statistics is
really unknown. The problem with loading the system with entities is that we don’t know where the loading
should take place and furthermore, we don’t know that our loading really hastens the transition to steady-state.
Finally the problem with deleting statistics is that we delete “good” as well as “bad” statistics and simulation
observations are important usually somewhat hard to obtain. So let’s consider a more thorough examination of
initialization bias in the context of the work cell problem.

Step 1: Open up the simulation from Chapter 5, section 6. For purposes of comparison, run the model
interactively for 40 hours.

Question 4: What is the average number of parts of type three in the system during the 40 hours?
________________________________________________________________________________

Simulation Output Analysis 173

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 5: What are the average time parts of type three are in the system during the 40 hours?
________________________________________________________________________________

Step 2: Add a SIMIO Experiment, specifying 10 replications, each replication being 40 hours. We will use
these results also for comparison.

Question 6: What is the 95% confidence interval on the average number of parts of type three in the
system during the 40 hours?
________________________________________________________________________________

Question 7: What is the 95% confidence interval on the average time parts of type three are in the system
during the 40 hours?
________________________________________________________________________________

Question 8: Do the averages from a single replication fall with the confidence intervals?
________________________________________________________________________________

Question 9: Why wouldn’t you be concerned if they didn’t?

________________________________________________________________________________

Step 3: Next, let’s add some plots of the two user-statistics so we can visualize the change in these over
time due to the initialization bias. Keep in mind this simulation starts empty and idle. Add two Animation
Plots from the Animation tab: one for the number of parts of type three in the system and the second for
the time parts of type three are in the system. Figure 11.1 shows the specifications for the average
number of part type three in the system. Figure 11.2 shows the specifications for the average time part
type 3 are in the system

Figure 11.1: Animation Plot for Average Number of Part type 3 in System

.
Figure 11.2: Animation Plot for Average Time Part type 3 in System
Step 4: Be sure that the Time Range on the plots is 100 hours, so we have enough time to see the changes.

Step 5: Change the run length of the simulation to 100 hours and run the simulation interactively (set the
speed factor to 1000) and observe the plots.

174 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 11.3: Average Number and Time in System
Step 6: Notice how the number in system plot changes rather markedly during the first 50 hours, while the
time in system plot doesn’t change much.

Question 10: During the initial 50 hours, how long does the transient period seem to last (recall this is the
time during which the behavior is erratic)?
________________________________________________________________________________

Question 11: After the transient period, there is a period of transition as the state of the statistics moves
from being highly unstable to becoming stable. How long do you think this transition period lasts?
________________________________________________________________________________

Question 12: How are the animation plots changing during the last 50 hours?
________________________________________________________________________________

When these plots stabilize, we say the statistics are in steady-state. Notice that we talk about the statistics, not
the system because it’s the statistics that interest us, even though it’s the system that is changing.

Question 13: Which statistic enters steady-state first? Number in system or time in system?
________________________________________________________________________________

Question 14: Since we can specify only one “warmup” period, what do you suggest?
________________________________________________________________________________

Step 7: We have seen only one replication during the 100 hour simulation. Under the Advanced Options
within the Run Setup section of the Run tab, click on Randomness and select the Replication Number
option. Set the replication number to “2” and re-run the 100 hour simulation. Now the time in system
continues to be relative stable, while the number in system shows a different transient behavior. Figure
11.4 shows the average number of part type three’s in the system during the first 100 hours.

Simulation Output Analysis 175

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 11.4: Replication 2 Average Number Part 3 in System
Question 15: How does the transient phase in Replication 2 for the average number of part 3 in the system
differ from the one for Replication 1 (which is shown in Figure 11.3)?
________________________________________________________________________________

Step 9: Now continue to change the replication number and interactively simulate, observing the behavior of
the average number of part 3’s in the system during the first 50 hours. Figure 11.5 shows the results for
replications 3, 4, and 5. We focus on number in system, but be sure to look at the time in system also.

Figure 11.5: Replications 3, 4, and 5 Average Number of Part Type 3’s in System
Step 10: With these five replications, we can see a range of behaviors at the beginning of the simulation. In
all cases the statistics appear converge to a common steady-state value, but there was a consistent transient
period during which there was no consistent behavior

176 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 11: Now simulate Replication 5 for 40 hours interactively and compare to Step 1.

Question 16: What is the average number of parts of type three in the system during the 40 hours?
________________________________________________________________________________

Question 17: What are the average time parts of type three are in the system during the 40 hours?
________________________________________________________________________________

Question 18: In comparing 40 hours with 100 hours simulations, do you think the differences are
important?
________________________________________________________________________________

Step 12: While the differences are probably not important in this case, you can see that they may be in other
cases. In any case we our simulation would benefit by the elimination of the initialization bias. Now the
question is how large should the warm-up be? Clearly we want to eliminate the transient period, which
after observing five different replications, lasts for about 20 hours. The transition period still shows some
effect, but we don’t want to throw away observations unless we think they are really bad. So let’s make
the warmup period 15 hours.

Replication Length

Now, as a rule of thumb, we do not want the warmup period to exceed 10-20% of the length of a replication.
In fact, if time permits, 5% would be a target.

Step 13: So let’s run the simulation experiment with a run length of 150 hours with a warmup of 15 hours
with a total of 10 replications.

Question 19: What is the 95% confidence interval on the average number of parts of type three in the
system?
________________________________________________________________________________

Question 20: What is the 95% confidence interval on the average time parts of type three are in the system?
________________________________________________________________________________

Question 21: How do these confidence intervals compare to the ones from Step 2 previously?
________________________________________________________________________________

In summary, it’s important to remove the initialization bias by using the warmup period. However you should
look at several replications of the warmup period in order to understand the range of behavior. Also you need
to examine all the statistics of interest, so your warmup covers the largest transient behavior. With the
warmup period established, be sure that the warmup period is less than 20% of the replication length. It is
generally a good rule to multiply the warmup period by 10 to obtain the replication length. Use a multiplier by
20, if your model runs quickly enough.

Part 11.4: Automatic Batching of Output

To obtain our previous confidence interval, we simulated (150 hours * 12 replications) for a total of 1800
hours. Of that total we used 450 hours for the warmup. Although this small problem runs through that
simulation time rather quickly, a larger problem may take hours of computer time to provide such a confidence
interval. So we may be interested in obtaining a confidence interval in less time. We “wasted” lots of
observations during warmup and we created fairly lengthy simulation replications.

An alternative to using replications is to employ a method known as “batch means”. Using batch means, we
attempt to create IID observations within one long replication. Since only one run of the simulation is used,

Simulation Output Analysis 177

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 we only lose the warmup observations once and we are able to invest our simulation effort in one long run,
rather than several. Hopefully, all this means that we save simulation time. The batch means method creates
“batches” of observations for which the average for the batches satisfy our IID requirement for requirements.
For example, the time-persistent statistics (state statistics), like number in the system may have a mean
computed every 25 hours if the means from every 25 hours appear to be IID. For observations-based statistics
(tally statistics), we might compute an average waiting time from each 50 customers, if the means from each
50 customers appear to be IID. The 25 hours for state statistics and the 50 customers for tally statistics are
called the “batch size”. There will be two batch sizes: one for state statistics and one for tally statistics. The
number of batches created during the simulation corresponds to the number of observations when computing a
confidence interval.

SIMIO will automatically employ a version of batch means to compute a confidence interval when: (1) user
statistics are defined, and (2) an SIMIO experiment with one replication is employed. The only SIMIO-
defined statistic that will be automatically batched are the sink node statistics.

Step 1: Use the simulation model from the previous section. Change the number of required replications to
one. Leave the warm-up at 15 hours and leave the replication length at 150 hours.

Step 2: When you execute the simulation, you will see a new row, called “HalfWidth” in the two user-
defined statistics for the average number of part type 3 in the system and the average time part type 3 are
in the system. You will also see that new row in the SnkPartsLeave statistics on flowtime. The output
for the two user-defined statistics are shown in Figure 11.6.

HalfWidth row

HalfWidth row

Figure 11.6: Output for User-Defined Statistics

Step 3: In the case of the time in system we see an “Insufficient” meaning we have not run the simulation
long enough for SIMIO to obtain enough batches to compute a confidence interval. It is also possible to
get a “Correlated” message meaning that the obtained batches do not satisfy the IID requirement. Figure
11.7 is an example of obtaining this message.

Figure 11.7: Example of Correlated Batch Means

Question 22: Why do the columns for “Average”, “Minimum”, and “Maximum” all have the same value?
________________________________________________________________________________

Question 23: Why does the “HalfWidth” show “NaN” (Not a Number)? Remember we are using only one
replication.
________________________________________________________________________________
178 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 4: Now change the replication length to 300 hours which is twice the length of the replication used
earlier.

Question 24: Do you obtain any “Insufficient” or “Correlated” messages on the statistics?
________________________________________________________________________________

Question 25: What is the 95% confidence interval on the average number of parts of type 3 in the system?
________________________________________________________________________________

Question 26: What is the 95% confidence interval on the average time parts of type 3 are in the system?
________________________________________________________________________________

Question 27: How do these confidence intervals compare to the ones from Section 3 previously?
________________________________________________________________________________

The automatic batching of statistics by SIMIO was able to produce a “legitimate” confidence interval with
only 300 hours of simulation, as opposed to the 1800 hours previously. In using the automatic batching, we
still needed the warmup analysis.

Part 11.5: Algorithms used in SIMIO Batch Means Method

SIMIO automatically attempts to compute a confidence interval via a batch means method for only the user-
defined tally and state statistics (the only exception is that the batch means method is also applied to sink
statistics) for a SIMIO Experiment having only one replication. It displays the results in the output in the row
called "HalfWidth", which is one half the confidence interval. It can fail to make the computation if it believes
there is insufficient data or if the batched data remains correlated. SIMIO does not include data obtained
during the Warm-up Period specified in the Experiment.

SIMIO’s batching method is done "on the fly," meaning the batches are formed during the
simulation. Thus batches can be enlarged, but not re-batched.

Minimum Sufficient Data

· For a Tally statistic, SIMIO requires a minimum of 320 observations
· For a State statistic, there must be at least 320 changes in the variable during at least 5
units of simulated time.
· For statistics that do not satisfy this minimum requirement, the half width is labeled
"Insufficient".
Forming Batches
· Form 20 batch means
o A Tally batch is the mean of 16 consecutive observations
o A State batch will be the time average over 0.25 time units
· Continue forming batches until there are 40 batches
o At 40 batches, the batch size is doubled (each batch is "twice as big"), and the number
of batches is reduced to 20
· Continue this procedure as long as the simulation runs. Notice that the number of batches
will be between 20 and 39

Simulation Output Analysis 179

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Test for Correlation
· SIMIO uses the VonNeuman's Test151 on the final set of batches to determine if the
assumption of independence can be justified.
o If the test fails, the half width is labeled "Correlated".
o If the test passes, the half width is computed and displayed.
Cautions
The SIMIO batch means method is only for steady-state analysis – it should be ignored completely
for terminating systems. There is no automatic determination of warmup period. All rules for
collecting batches are somewhat arbitrary, so some care needs to be exercised in using it.

Part 11.6: Input Analysis

It may seem strange to include “input analysis” in a chapter devoted to “output analysis” but it must be
remembered that the analysis of the input is based on the behavior of the output. SIMIO provides two input
data analysis: (1) Response Sensitivity Analysis and (2) Sample Size Error Analysis. Both analyses require
that at least one of your inputs to be an Input Parameter and that you have defined at least one Experiment
Response. Although we briefly introduced response sensitivity analysis previously, we will describe it as well
as the sample size error analysis within the same context, namely the workcell problem used throughout this
chapter.

Recall that “input parameters” are defined from the “Data” window tab in three possible ways: (1) through a
DISTRIBUTION you hypothesized, (2) with a TABLE of observations that you have collected, and (3) using a
SIMIO EXPRESSION. Any defined input parameter can be used anywhere in your model and various objects
can reference the same input parameter. Using distributions have the advantage that SIMIO will provide you
with a histogram based on 10,000 samples to show you what that distribution looks like.

Step 1: Save the current model and delete the animation plots as they will not be needed.

Step 2: Rename the model state variable to GStaNumberInSystem. Modify the two user-defined
statistics so we have some possible responses for experimentation. Call the state statistic
StateStatNumberInSystem and the tally statistic TallyStatTimeInSystem. Be sure that the state statistic
references the GStaNumberInSystem state variable and that the tally statistic has Time as the Unit
Type.

Step 3: Fix the State Assignment at the SrcParts and at the SnkPartsLeave so that
GStaNumberInSystem is always incremented (add one) at the source and decremented (subtract one) at
the sink. Modify the Tally Statistics collection at the Input@SnkPartsLeave so that it corresponds to
Figure 11.8.

Figure 11.8: Tally All Entities Time in System

Step 5: Next, we need to create the input parameters, which will correspond to, in this case, all the
distributions used in the model. These are created by specifying a “Distribution” with the Input

151
This is a statistical test of independence.
180 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Parameters section from the “Data” Tab. Make sure to change the Unit Type property to “Time” and
specify minutes. Figure 11.9 shows the inputs and these are the same the SequencePart table.

Figure 11.9: Definition of Input Parameters

Step 6: Select the SrcParts and use the InpDistPartsArrive as the Interarrival Time.

Step 7: Also substitute the appropriate input distribution into the SequencePart Data Table for each of the
ProcessingTimes values as seen in the partial view of the table in Figure 11.10.

:
Figure 11.10: Partial View of the Sequence Part Table using the New Input Distributions
Step 8: Next select the Experiment and add two experiment responses as shown in Figure 11.11.

Figure 11.11: Experiment Responses for Number and Time in System

Step 9: Run the experiment with a Warm-up Period of 15 hours and a run length of 150 hours with 20
replications which will be used as a base reference for the results.

Question 28: What did you get for the 95% confidence interval on the average number in the system?
________________________________________________________________________________

Question 29: What did you get for the 95% confidence interval for the average time in the system?
________________________________________________________________________________

Simulation Output Analysis 181

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Response Sensitivity Analysis

The response sensitivity analysis measures how each response is affected by changes in the input parameters.
It will be computed for each scenario in the experiment. A linear regression relates each input parameter to
each response, so it is necessary that the number of replications be greater than the number of input
parameters.

Step 10: Our prior model has ten input parameters and we are using 20 replications. Select the Input
Analysis tab and then the Response Sensitivity icon and run the model (it may be necessary to Reset it
first). The results of the sensitivity analyses are shown in a Tornado Chart, a Bar Chart, and a Pie Chart.
The raw data for each of the regressions are also available.

Figure 11.12: Tornado Chart for Time in System

Step 11: Refer to the Tornado Chart in Figure 11.12 and pass your cursor over the bars. If an input
parameter has a negative coefficient, then when it is increased, it reduces the response. If the input
parameter has a positive coefficient, then when it is increased, it increases the response. The magnitude of
the coefficient determines it’s important relative to the other input parameters.

Question 30: What input parameter will reduce the time in system the most when its distribution is
increased?
________________________________________________________________________________

Question 31: What input parameter will increase the time in system the most when its distribution is
increased?
________________________________________________________________________________

Question 32: What input parameter reduces the number in system the most when its distribution is
increased?
________________________________________________________________________________

Question 33: What input parameter increases the number in system the most when its distribution is
increased?
________________________________________________________________________________

Question 34: What surprises you about these results?

________________________________________________________________________________

182 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 35: What is the primary value of the pie chart relative to the sensitivity analysis?
________________________________________________________________________________

Sample Size Error Analysis

The sample size error analysis attempts to assess the uncertainty in the responses relative to the uncertainty in
the experiment and the uncertainty in the input parameter data. To perform this analysis, it is assumed that the
input parameters were estimated from collected data samples. In other words you actually have a record of
observations that you used for a given input parameter estimation (or distribution fitting).

Step 12: Pretend now you collected data for the input parameter estimation. From the Mode and the Input
Parameters on the “Data” tab. Select a defined input parameter. Change the Number of Data Samples
property according to Table 11.1 and be sure that the Include Sample Size Error Analysis property is True.

Table 11.1: Data Samples for Each Input Parameter

Input Parameter Data Samples
DistPartsArrive 125
DistPart1StationA 85
DistPart1StationC 85
DistPart2StationA 100
DistPart2StationB 100
DistPart2StationC 100
DistPart3StationA 60
DistPart3StationB 60
DistPart4StationB 45
DistPart4StationC 45
Step 13: Go to the Input Analysis tab within the EXPERIMENT. Select the Sample Size Error icon and run the
analysis (after doing Reset). This may take a little bit of time to complete.

Step 14: Next run the experiment from the design view to obtain a SMORE plot.

Step 15: The two important tabs are: (1) Contribution to Uncertainty and (2) Benefit of Additional Samples.
Look first at the contribution to uncertainty, shown in Figure 11.13.

Simulation Output Analysis 183

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 11.13: Uncertainty in Time In System
Question 36: Which factor has the biggest impact on time in the system?
________________________________________________________________________________

Question 37: The SMORE plot illustrates the confidence interval for the mean (in tan), with the expanded
half width due to input uncertainty in blue (and stated in text). How much does the uncertainty in the
input parameters “expand” the output uncertainty?
________________________________________________________________________________

Question 38: Which uncertainty then, input or output, dominates the overall uncertainty in this scenario?
________________________________________________________________________________

Step 16: Next look at the tab: Benefit of Additional Samples as shown in Figure 11.1.

Figure 11.14: Benefit of Additional Samples

184 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 39: Which input parameter would benefit the most from collecting additional observations for
time in the system?
________________________________________________________________________________

Step 17: Go back to the Input Parameters and change the Number of Data Samples property for
InpDistPartsArrive to 250. Next repeat the Sample Size Error analysis again.

Question 40: How does the Contribution to Uncertainty change?

________________________________________________________________________________

Question 41: What happens to the Benefit of Additional Samples?

________________________________________________________________________________

Question 42: Would you now put more time into observing InpDistPartsArrive or would you allocate
observations to other input parameters
________________________________________________________________________________

Part 11.7: Commentary

· Sometimes the modeling effort consumes so much of a simulation project that the analysis of the model
is almost overlooked. In particular, it is easy to forget that simulation output can be difficult to interpret
and requires time and care. The issue of bias is often ignored, but can be a critical mistake.

· Replications permit the computation of confidence intervals and confidence intervals are our best way to
measure of mean and variance of the expected value.

· Our experience is that most simulations can be run quickly, which means that elimination of the warmup
and the use of replications becomes feasible even for steady-state simulations. However there are some
steady-state simulations that take excessive computer time and so the idea of using batch means
becomes attractive.

· The availability input analysis features in SIMIO should encourage you to consider the importance of
the simulation inputs relative to your performance measures. You can judge the value of the various
inputs and consider the amount of data collection devoted to their specification.

Simulation Output Analysis 185

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Chapter 12
Materials Handling
In the STANDARD LIBRARY, You have already noticed that there is a VEHICLE object and a CONVEYOR path.
Perhaps the term “vehicle” already conjures up the idea of moving materials or people from place to place and
your intuition would be correct. The VEHICLE object can pick up and delivery entities which can be used to act
as a fork lift truck in a manufacturing operation, a human stock handler in a warehouse, a taxi cab for picking
up and delivering people as well as a bus, a train, or other type of people mover to move entities either
individually or in a group. The vehicle is an individual object whose routing is based on either “demand” or
on a “fixed path”. A conveyor in SIMIO, on the other hand, provides a continuous platform for conveying
items or people. It can represent a wide range of industrial devises like gravity conveyors, belt conveyors,
powered overhead conveyor, and even a power and free conveyor. Often materials handling in industry will
employ a variety of materials handling devises. A distinction among conveyors used by SIMIO is whether the
conveyor “accumulates” or doesn’t (“non-accumulating). An accumulating conveyor allows entities to queue
up at its end. A non-accumulating conveyor will stop when an entity gets to the end and won’t start up until
that entity is removed.

A key to modeling materials handling problems in SIMIO is the use of the TRANSFERNODE and to somewhat a
lesser extent, the external, OUTPUTBUFFER which are attached to many of the objects like the SOURCE and the
SERVER. The TRANSFERNODE (unlike the BASICNODE) provides the “Transport Logic” which means you can
request a vehicle to move the entity to its next designation. If that transport device is not available, then the
entity must wait. The TRANSFERNODE can be used alone and entities can wait there for pickup at that node but
they will not queue up in the usual fashion. For other objects, the OUTPUTBUFFER provides a place for the
entities to wait for transportation.

Part 12.1: Vehicles: Cart Transfer in Manufacturing Cell

We will reconsider the original problem of Chapter 5, whose layout and information are shown in Figure 12.1
Recall there are four part types, each with its own source and each part type has its own routing through the
manufacturing cell. However, we now learn that instead of leaving the system, parts actually have to be put on
a cart and wheeled to the warehouse building 300 yards away (see Figure 5.9). The cart can carry a maximum
of three parts and travels at four feet per second. After the cart has finished dropping off parts, it will return to
the pickup station for the next set of parts or wait there until a part is ready to be taken. The cart does not have
to be filled for it to be taken to the warehouse (i.e., will carry one part if it is available).

Step 1: Open up the last model from Chapter 5 and save it as chap12.1.

Step 2: To facilitate a cart taking the parts to the warehouse, delete the CONNECTOR between the SINK
object and the prior BASICNODE. Next, insert a TRANSFERNODE in between the SnkPartsLeave named
TNodeCart as seen in Figure 12.1.

Step 3: Select the Input@SnkPartsLeave and remove the Tally Statistics as VEHICLES will be entering
instead. To calculate the same statistics, we could insert a Tally step in the Entered add-on process
trigger.

Step 4: Connect the BASICNODE to the TRANSFERNODE with a PATH with a logical length of 10 yards.
Finally, connect the TNodeCart TRANSFERNODE to the SINK with a PATH making sure it’s Type is set to
“Bidirectional” so our cart can move in both directions. Set the path’s logical length to 300 yards.

186 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 One Machine Two Machines
(2, 5, 8) Minutes (5, 9,11) Minutes

Station A Station B Cart parking space

Parts have different 25 yds 35 yds 20 yds

part routings
Parks
P at Home Transfer
Parts 10 yds Node when Idle
Arrive
40 yds 30 yds
Parts move at 2 mph

Station C
Carry a Max of 3 Parts
Two Machines First 4 hours
One Machine Second 4 hours 300 yds
(2, 6, 11) Minutes

Cart Travels at
4 feet per second

Parts
Leave

Figure 12.1: Adding a Vehicle to Carry Part to the Exit

Step 5: Drag and drop a VEHICLE object from the [Standard Library] to the modeling canvas in the facility
window named VehCart similar to Figure 12.2. Make sure to increase the size of the RIDESTATION queue
which animates the parts being transported so it can hold three parts.

TNodeCart
TRANSFERNODE

Create space for

up to three parts to
be carried by
stretching the
cart’s ride queue

Figure 12.2: Adding a Vehicle

Step 6: Change the TNodeCart TRANSFERNODE’s Ride On Transporter property to “True” and specify the
“VehCart” as the Vehicle Name which was just added (see Figure 12.3). Every time an entity passes
through this node, it will now generate a ride request to be sent to the vehicle. The entity cannot continue
until the vehicle is there to pick it up.

Materials Handling 187

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 12.3: Setting up the Node to Have Entities Utilize a Vehicle
Step 7: Use Figure 12.4 as a guide to set the VehCart’s Initial Desired Speed to four feet per second and
Ride Capacity to three. Also, make sure to set the Initial Node(Home) to the TNodeCart node which was
just added, as this will be our vehicle’s home node. Leave the Routing Type specified as “On Demand”,
meaning it will respond to requests for transportation from the stations as needed. Now change the
VehCart’s Idle Action property to “Park At Home” so that the vehicle will always go to the home node if
there are no entities with ride requests.152 Otherwise the VEHICLE would remain at the SINK until a ride
request is generated.

Figure 12.4: Vehicle Properties

Step 8: Run the simulation now to see how the cart works in the animation. As you can see the VehCart
originally is parked at the TNodeCart node as seen in Figure 12.5. By default, the parking animation
queue is above the node153 and oriented from left to right. If there was more than one vehicle they would
be stacked on top of one another again facing left to right in the facility as seen in the right picture of
Figure 12.5 where there are three carts in the system.

152
The Network Turnaround Method can be changed for example to “Reverse” so the cart would go forward to the exit and then look
as if it backs up to the TNodeCart node. By default it will turn around and move forward back to the home TNodeCart node.
153
Note the familiar green line that represents the animated queue line is not visible in this case.
188 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 12.5: Vehicle is parked at the Home Node using Default Parking Station
Question 1: Does the Cart pick up more than one part?
________________________________________________________________________________

Question 2: What do you notice about how a single part or two parts ride on the vehicle?
________________________________________________________________________________

Step 9: To fix the issue of a single part riding out to one side, select the RIDE STATION queue on the
VEHICLE and change the Alignment from “None” to “Point” under the Appearance tab. Add an additional
vertex to the queue to accommodate three parts and move the first orientation point which is on the bottom
to the middle as seen in Figure 12.6.

Last Part Rides

in Middle

Figure 12.6: Changing the Orientation of the Riding Queue

Step 10: Run the simulation now to see how the cart works in the animation.

Step 11: By default, a parking location for a node is automatically added above the node as seen in Figure
12.7. However, you should generally not utilize the default one since the location and/or shape of the
parting station cannot be adjusted as well as the orientation of the VEHICLES. Select the TNodeCart node
by clicking it and then toggle the Parking Queue option in the Appearance tab to be clear. Next, a parking
location is needed for the vehicle or the vehicle will vanish when parked.154 Click on the TNodeCart node
and select “Draw Queue” from the Appearance Tab. Click on the ParkingStation.Contents. A
cross cursor will now appear in the facility window. Left click where you want the front of the queue to be
and then right click where you want it to end (before you right click to end the queue you could also
continue left clicking to add more vertices). Place it above the node so the Cart will be facing downward
as seen in Figure 12.7. Click on the queue and in the properties window make sure the Queue State is
ParkingStation.Contents. This queue will animate the parking station of the TNodeCart
TRANSFERNODE for vehicles parked at the node. It can be oriented and placed at any location. Figure 12.7
shows the same scenario as Figure 12.5 but this time we are able to orient the vehicles to pointing in the
same direction as the path.

154
The parking queue is again just for the animation since the vehicle is physically at the node even though it cannot be seen.
Materials Handling 189

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Detached Queue
for Parking Three Carts
Station Waiting in Line

Figure 12.7: Vehicle is parked at the Home Node using the New Station
Step 12: Run the simulation now to see how the cart works in the animation.

The model does not show how many parts are waiting for the cart at the transfer node so it may be helpful to
see how many are there at any given time in the simulation. To accomplish this we can add an animated
“Detached Queue” for the TNodeCart.

Step 13: Click on the TNodeCart node and select “Draw Queue” from the “Appearance” Tab. Click on the
“RidePickupQueue”.155 A cursor will now appear in the facility window. Left click where you want
the front of the queue to be and then right click where you want it to end (before you right click to end the
queue you could also continue left clicking to add more vertices). As seen in Figure 12.8, arrange this
detached queue in an “L” shape. Otherwise an entity is shown trying to enter the transfer node will be
visible156 and by overlaying the “L” it shows all the entities waiting for transportation.157

Detached Queue
used for ride
pickup entities

Figure 12.8: Ride Pickup Queue

Step 14: Run the simulation to see how the DETACHED QUEUE works.

Question 3: What is the utilization of the VehCart during a 40 hour replication?

________________________________________________________________________________

155 If added just generic animated queue now, change the Queue State property to TNodeCart.RidePickupQueue where TNodeCart
is the name of the transfer node where the parts are being picked up. Your transfer node’s name may be different.
156
You can actually see the entity attempting to enter the transfer node if you watch that node carefully.
157 These queues are for animation purposes only. The parts physically reside at the entrance of the node and one will always be visible
as the others will stack up underneath. In later chapters, the notion of a STATION to hold entities will be discussed.
190 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Part 12.2: Cart Transfer among Stations

Now, suppose that the transportation of parts among the stations, including entry and exit are handled by two
similar vehicles. These vehicles are different than the VehCart which carries the parts to the warehouse SINK.
These carts will wait at the beginning of the circular network until they are needed.

Step 1: Convert the one BASICNODE that connects the four SOURCES into a TRANSFERNODE by right clicking
and selecting the Convert to Type menu item. This will be necessary in order for parts entering the
network from the sources the ability to request transportation on the inside carts. Next, name this new
TRANSFERNODE TNodeStart.

TNodeStart
TRANSFERNODE

Figure 12.9: Changing the Start Node to a TRANSFERNODE

Step 2: Now add the inside carts by inserting another VEHICLE from the [Standard Library] into the model
naming it VehInsideCart. Shrink the vehicle leaving the ride station queue in the same position.

Step 3: Make the VehInsideCart have an initial population of two158 which will allow you to see the
simultaneous behavior of multiple vehicles on the same path. The new vehicle will have the
characteristics, as shown in Figure 12.10 (i.e., the Initial Node should be TNodeStart and the Idle Action
property should be set to “Park at Home”). Select the “Transporting State” from the “Additional Symbols”
and color it orange.

Notice that the VEHICLE object is similar to the ENTITY object in that multiple instances, which we call “run-
time” instances can be created from the “design-time” instance of the VehInsideCart .159 “Design-time”
occurs while you are putting instances of the objects onto the modeling. All the objects in the FACILITY
window are initialized just before the simulation begins to execute. “Run-time” occurs when the simulation
executes.

158
The fact that the property description says Initial number in System should indicate that vehicles can be added and removed from
the simulation during the simulation execution (something that is left to a later chapter).
159
We will see that the WORKER object can also have run-time instances.
Materials Handling 191

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Orange Color of
InsideCarts Specify “home” as
TNodeStart

All carts park at

home when idle

Specify two
Inside Carts

Figure 12.10: Properties of the InsideCart

Step 4: Now we need to make sure that the inside carts can drop off parts and then return back to the
circular path. Since they obviously cannot go through the SERVERS, the inside carts would move to the
input node of a SERVER, drop off the part and then become deadlocked (i.e., stuck). 160 The easiest
solution is to allow the carts to maneuver around the stations (i.e., add a CONNECTOR from the input to the
output nodes of each station as shown in Figure 12.11 making sure the direction of the link is correct.161
Connector from
Input to Output

Figure 12.11: Adding a Connector at a Station

Step 5: Next, select the TNodeStart and the three output nodes (i.e., TRANSFERNODEs) of the station
servers and specify the entities leaving these nodes to ride on an VehInsideCart as seen in Figure 12.12.

160
By default, WORKERS and VEHICLES will not enter fixed objects like SERVERS, SINKS, COMBINERS, etc. One solution is to create a
connector leaving the input node of the server to the transfer node as well as another one from the transfer node to the output node to
allow pickup.
161Recall, CONNECTORS take up zero time as well as have zero length in the model. Basically, the inputs and outputs of the server are
really the transfer nodes of the circular path.
192 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 12.12: Use the Inside Cart
Step 6: Finally change the path between the exit node on the inside loop and the TNodeCart node to two
one-way paths (i.e., add an additional path from the TNodeCart back to the basic node making sure to
make the path ten yards long). Also change the path between the TNodeCart and the SnkPartsLeave to
two one-way paths as well. Now your overall model should look like the one in Figure 12.13.162

Two Paths

Figure 12.13: Model with Vehicles

Step 7: Also, change the VehCart Population property Initial Number in the System to “2” so two carts
will be available for the outside transportation. The TNodeCart node should also be specified as shown in
Figure 12.14.

Figure 12.14: Specifying TNodeCart

Step 8: Save and run the model observing what happens.

Question 4: What did you notice about the utilization of the outside carts?
________________________________________________________________________________

162One can use bidirectional paths but this can lead to bottlenecks and deadlock as vehicles, workers, and entities can only be going in
one direction and must wait for all items to clear path before the direction can be reversed. It is always better to use to separate paths.
Materials Handling 193

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 9: The difficulty was the parts were not unloaded from the inside cart to be transported by one of the
two outside carts to the SINK. The issue is once the part has finished its last processing on a station, the
next sequence is the exit. Therefore, the cart takes the part to its next destination (i.e.,
Input@SnkPartsLeave). To force the inside cart to drop off the parts at the TNodeCart node, this node
should be inserted into the four part sequences right before the Input@SnkPartsLeave row for each type
as seen in Figure 12.15.163

Figure 12.15: Adding the Transfer Cart to Force a Drop off of the Parts
Step 10: Once all four sequences have been modified, save and execute the model. Observe the interacting
behavior of the carts as well as notice how the carts park.

Question 5: How do parts get transported from the inside workstations to the outside cart for transfer to the
sink where the parts leave?
________________________________________________________________________________

Question 6: Is the queue ride parking for the TransferCart node needed anymore?
________________________________________________________________________________

Question 7: What is the utilization of both (outside) Carts as well as the utilization of both InsideCarts at
the end of the simulation (24 hours)?
________________________________________________________________________________

Question 8: How are the individual vehicles within a cart type distinguished in the output from each other?
__________________________________________________________________________________
______________________________________________________________________________

Step 11: There are a few anomalies that occur within the model. Notice how the inside carts will wait while
they drop off a part when an outside cart is not available. When the part starts the drop off process, it visits
the node which informs the part that it requires a vehicle to continue and therefore never completes the
drop off process which releases the inside cart. To fix this anomaly along with other animation issues,
delete the TNodeCart node and insert a new SERVER named SrvPickup with an infinite capacity and zero
for the processing time as seen in Figure 12.16.

163
The easiest way to insert the node into the sequence is to select the Input@SnkPartsLeave row and then click the Insert Row
button in the Data section of the Table tab.
194 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 12.16: Adding a new Server as Pickup Point
Step 12: Reconnect the two 10 yard paths from the path network to the input of the SrvPickup and the two
300 yard paths from the output node to the SINK.

Step 13: For the sequence table, change the TransferCart node to the Input@SrvPickup as well as make the
home node for the Cart the Output@SrvPickup node as seen in Figure 12.17.

Figure 12.17: Changing the Sequence Table and Cart Home Node
Step 14: Finally, request for transportation on the Output@SrvPickup transfer node using the same
information from Figure 12.14.

Step 15: Save and run the model observing what happens with the two types of carts.

Question 9: Now what is the utilization of both (outside) Carts as well as the utilization of both
InsideCarts at the end of the simulation (24 hours)?
_____________________________________________________________________________

Part 12.3: Other Vehicle Travel Behaviors (Fixed Route and Free Space Travel)

Now let’s experiment with different vehicle specifications to create different cart behaviors. We will now
explore both a fixed route travel (i.e., bus) as well as travel through free space.

Materials Handling 195

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Fixed Route Travel

Currently the VehInsideCart travels to wherever transport is needed among the stations including entry and
pickup. That Routing Type, specified on the vehicle definition, is called “On Demand.” Suppose now we
want the cart to travel in a fixed sequence around all the service points in the model.

Step 1: From the Data tab, insert a SEQUENCE TABLE for the vehicle Route Sequence named SeqInsideCart as
shown in Figure 12.18. Note the sequence includes both input and output nodes of the stations.

Figure 12.18: New Route Sequence for InsideCart

Step 2: Select the VehInsideCart and set the Routing Type property to “Fixed Route” and the Route
Sequence to the SeqInsideCart sequence table as seen in the right picture of Figure 12.18.

Step 3: Run the model and observe both the behavior of the cart(s) and the statistics on the cart performance.

Question 10: Are both inside carts behaving as expected? What is their utilization?
________________________________________________________________________________

Question 11: Can you see the second cart? Why is it’s utilization 100%?
________________________________________________________________________________

Step 4: Because both carts are always active, their utilization is 100% and the two carts ride (i.e., overlap
one another) in the same location. So if we are to see carts we need to interrupt the constant motion or
have the two carts start at different times (i.e., have a Source create them at different times). There several
ways to achieve the result.

Step 5: Change the Load Time of the VehInsideCart to an Exponential(0.5) minutes. Save and run the
model observing the cart behavior.

Question 12: What happens to both inside carts?

________________________________________________________________________________

Free Space Travel

Rather than travel on a specific path between stations, vehicles, workers and entities can travel in “free space.”
To do this, SIMIO needs the “location” of the various nodes. In other words, it’s going to assume that your
model is “to scale” as it might be if the animation layout is based on computer-aided drawing (CAD) software.
It’s important to recognize that the time it takes parts to be moved by vehicles depends on the distance the
vehicle travels and its travel speed164. To illustrate the free space travel, let’s treat our animation as though it
came from a CAD drawing.

164 When vehicles move entities, it’s the speed of the vehicle that determines how fast it traverses a distance and not the entity speed.
196 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 6: Save the current model as Chap12.3Step6 and eliminate all the paths and nodes between the
stations and the SrvPickup so that travel is unrestricted, as seen in Figure 12.19.

Figure 12.19: Layout to Scale: No Paths

Step 7: Modify the VehInsideCart “Travel Logic” and “Routing Logic” as Figure 12.20. In this case we
are showing there is no travel network – meaning the vehicle must use “Free Space”. We also are routing
On Demand and the Idle Action is to Remain in Place.

Figure 12.20: Travel Logic and Routing Logic

Step 8: Run the model and observe the travel of the inside carts.

Question 13: When would this type of travel be most appropriate and what do you observe about the
parking?
________________________________________________________________________________

Part 12.4: Conveyors: A Transfer Line

SIMIO provides for two kinds of conveyor modeling concepts: accumulating and non-accumulating. An
accumulating conveyor allows parts to queue at its end, but the conveyor keeps moving. This approach might
model a belt conveyor whose parts rest on “slip sheets” that allows the belt to continue to move. A non-
accumulating conveyor must stop until the on-off operation completes. This concept might apply to an
overhead conveyor with parts on carriers. SIMIO conveyors do not have on/off stations, so that a transfer line
will have to be modeled as a series of individual conveyors. Since the transfer line stops, the non-
accumulating concept applies. However the individual conveyors will need to be synchronized.

Materials Handling 197

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Suppose now that the transfer among stations in the manufacturing cell is handled by a transfer line which is a
continuous conveyor, possibly in a loop, that has on/off stations. When a part is at an on/off station, the entire
line pauses until the part has been placed on or taken off the conveyor. It takes between 0.5 and 1.5 minutes to
load/unload the parts from the nodes.

Step 1: Reopen the last model from Chapter 5 and save it as a new name. Change the arrival rate of the
source to be Exponential(15) now.

Step 2: Change all the paths on the interior of the model to conveyors by selecting the path and accessing
the Convert To Type submenu via a right mouse click.165 To embellish the model, add path decorators for
conveyors as seen in Figure 12.21 from the Path Decorators tab.

Step 3: Also modify the symbols for the parts so they are one meter wide by one meter in length by one-
half meter in height. Select the EntParts MODELENTITY and for each Active Symbol, change the size under
the General→Physical Characteristics→Size properties. After selecting a different active symbol, you
will need to deselect it and the reselect it to see the size for that symbol. We will need to be a little more
specific about these sizes as we size the conveyors. We created these symbols from the “Project Home”
tab in the “Create” section, clicking on “New Symbol” and selecting “Create a New Symbol”. When
creating the symbol we paid close attention to its size (the grid is in meters).

Conveyors

New Part
Symbols

Figure 12.21: 3D View with Conveyors

Step 4: Next, properties of the conveyors need to be specified. Here it is convenient to use the
“Spreadsheet View” by right-clicking on one of the conveyors and selecting Open Properties Spreadsheet
View for all objects of this type. The lengths from the previous paths (recall they were 25, 35, 20, 30, and
40 yards) do not need to be modified. All five conveyors should have a conveyor speed of 0.5
meters/second for each conveyor. Set the Accumulating property to False for each conveyor (i.e., the
selection boxes should be unchecked).

Step 5: In the Entity Alignment property, choose the Cell Location option. Choosing this option causes the
conveyor to be conceptually divided into cells. Parts going onto the conveyor will need to be placed into a
cell. Therefore choose the number of cells to be equal to number associated with the conveyor length.
Otherwise, if you select Any Location for the Entity Alignment, a part can go onto the conveyor at any
place, even interfering with another.

165
Note you can use the Ctrl key to select all five paths and convert them all at once.
198 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 12.22: Specifying Properties for all Five Conveyors of the TransferLine (Spreadsheet View)
Step 6: Since the transfer line consists of multiple conveyors rather than a continuous one, the conveyors
have to be synchronized so that when parts enter an on/off location on the transfer line, the entire line
which is a set of conveyors stops to await the on/off operation.166 To do this, first define a new DISCRETE
STATE VARIABLE of type integer (from the Definitions tab) for the Model named
GStaNumberNodesWorking. This variable will represent the number of on/off locations that are
currently either putting parts onto or taking parts off the transfer line. When this state variable is zero, the
transfer line is running and when this variable is greater than one, the transfer line should be stopped.

Step 7: Let’s now insert a new “process.” From the “Processes” tab, select “Create Process” and name the
new process On_Off_Entered. The process will consist of the steps shown in Figure 12.23.

Figure 12.23: Process when Entering/Leaving the Transfer Line

· When a part reaches an on/off operation, the Assign step needs to increment the
GStaNumberNodesWorking by one and all conveyors need to be stopped by setting their desired
speed to zero (see Figure 12.24).167

Figure 12.24: Incrementing the Number Nodes Working and Shutting Down the Conveyors.
· The Delay step will be used to model the on/off time, which is assumed to be uniformly distributed
with a minimum of 0.5 and a maximum of 1.5 minutes.
· The second Assign step will now decrement the GStaNumberNodesWorking by one.
· The Decide step is based on the condition that GStaNumberNodesWorking==0, meaning no
other on/off stations are busy loading/unloading the transfer line.

166
The synchronization is not exact since the occupied cell of the upstream conveyor is not known to the downstream conveyor and
parts from a cell may conflict with parts arriving from the upstream conveyor (i.e., SrcParts).
167 You will need to select the Assignments (More) button in the Assign step.
Materials Handling 199

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 12.25: Decide Step Option
· The last Assign step will turn the conveyors back on. SIMIO stores all the times internally in meters
per hour. If the state variable has units these can be specified in the assignment. Make sure you change
the units to meters per second or an assignment of 0.5 would be treated as 0.5 meters per hour.168 You
can convert the original specification of 0.5 meters per second to 1800 meters per hour to be consistent
with how SIMIO will interpret the assignment and avoid having a conversion each time. Figure 12.26
shows the conveyors being turned back on.

Figure 12.26: Turning the Conveyors Back On

Step 8: Finally, the On_Off_Entered generic process needs to be invoked appropriately (i.e., every time a
part is loaded or unloaded to the transfer line). Select the On_Off_Entered as the add-on process trigger for
all the nodes and one path (i.e., the path to the exit SnkPartsLeave) as specified in Table 12.1.

Table 12.1: All the Add-On Process Triggers that Need to Go Through the On/Off Process Logic
Add-On Process Trigger Which Object the Trigger is Associated
Exited Output@SrcParts (TRANSFERNODE)
Entered Input@StationA (BASICNODE)
Exited Output@StationA (TRANSFERNODE)
Entered Input@StationB (BASICNODE)
Exited Output@StationB (TRANSFERNODE)
Entered Input@ SnkPartsLeave (BASICNODE)
Entered Input@StationC (BASICNODE)
Exited Output@StationC (TRANSFERNODE)

Question 14: Why did we increment and decrement GStaNumberNodesWorking by one rather than just
setting it to one or zero which would eliminate the need for the Decide step?
________________________________________________________________________________

Question 15: Why not simply use the On_Off_Entered process at the BASICNODE of the entry/exit?
________________________________________________________________________________

168
The presumption of a standard internal time probably represents a choice of efficiency over ease of use, since determining the
original time units would represent some additional computational effort.
200 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 9: Run the simulation for 40 hours at a slow speed to be sure each of the parts follows their sequence
appropriately and answer the following questions.

Question 16: How long are all part types in the system (enter to leave)?
________________________________________________________________________________

Question 17: What is the utilization of each of the SERVERS?

________________________________________________________________________________

Question 18: What is the average number on the Conveyor2 link?

________________________________________________________________________________

Question 19: What is the average time on the link for Conveyor3 link?
________________________________________________________________________________

Part 12.5: Machine Failures in the Cell

Failures are a part of most mechanical and other systems. SIMIO provides methods to model these failures
within the “Reliability Logic” section of the SERVER (and other objects). In the case of the transfer line, we
also need to shut down the line while a machine is in the state of being repaired.169 Let’s assume that
SrvStationC is the unreliable machine. In particular, we will assume that the MTBF (Mean Time Between
Failures) is exponentially distributed with a mean of ten hours, while the MTTR (Mean Time To Repair) is
modeled with a Triangular distribution (0.5, 1, 2) hours.

Step 1: In the “Reliability Logic” section for the Station C SERVER, specify the Failure Type to be
“Calendar Time Based”, the Uptime Behavior be Random.Exponential(10) with units of hours, and
the Time To Repair should be Random.Triangular(0.5,1,2)with units of hours as seen in Figure
12.27. Remember that the server is failing and that means that all the capacity associated with this server
fails (the capacity for this server is scheduled, but it changes between 1 and 2).

Figure 12.27: Specifying a Failure and Repair

Step 2: When the SrvStationC fails, the entire transfer line should be shut down and then started back up
once the station has been repaired. Select SrvStationC and create new Add-On Process Triggers,
“Repairing” and “Repaired” to produce new processes SrvStationC_Failed and SrvStationC_Repaired.
The “Failed” trigger will be invoked at start of the failure while the “Repaired” trigger will be invoked
after the SERVER has been repaired. In SrvStationC_Failed, use an Assign step to turn off all of the
conveyors as by setting their DesiredSpeed to zero. In the SrvStationC_Repaired, use an Assign
step to turn back on the conveyors if there is currently no nodes loading/unloading (i.e., setting the
DesiredSpeed = 0.5 meters per second for all conveyors) as seen in Figure 12.28.170

169 The tremendous cost of downtime due to everything on the line becoming idle is one of the reasons why transfer lines and similar

highly automated processes are becoming relatively rare.

170
Copy the Decide and Assign steps from the On_Off_Entered process and remove the increment of the
GStaNumberNodesWorking.
Materials Handling 201

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 12.28: Failed and Repairing Process Triggers
Step 3: Also, the On_Off_Entered process should not start up the transfer line if the SrvStationC has
failed and is currently being repaired. To do this change the Decide step in the On_Off_Entered
process so the conditional expression becomes the following.171

(GStaNumberNodesWorking==0) && (SrvStationC.Failure.Active==False)172

Step 4: Note that the SrvStationC will change to the color red when it is down, indicating that the server
(and all its capacity) is in the “Failed” state. Run the model and make sure the transfer line stops when the
station is also under repair.

Question 20: Is there anything about the final animation that troubles to you?
________________________________________________________________________________

Part 12.6: Sorting Conveyors

Now we find out that the output from the conveyor system is another system of conveyors which sorts the
parts into destinations. The parts that exit the manufacturing cell are routed to an accumulating conveyor at a
labeling machine. From the labeling machine, the parts are conveyed to one of four sinks corresponding to the
specific part type as seen in Figure 12.29. The four conveyors directly in front of the sinks are accumulating
conveyors. These conveyors keep moving and parts are allowed the queue, although in this case the sinks
pose no barrier to exit. The rest of the conveyors are non-accumulating.

Labeler
BasicNodes Sorting Conveyors

Accumulatin
g Conveyor

Figure 12.29: Labeling and Sorting Conveyors

171 Object.Failure.Active function will return true if the current object has failed.
172
Notice that SIMIO utilizes C# logical operators (i.e., == for equality, && for “And” statement, and | | for the “Or” statement).
202 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 1: Remove the original SnkPartsLeave and replace it with a SrvLabeling. The labeling processing
time follows a Triangular(1,2,3) minutes and has a capacity of “1.” Since the labeler will be supplied by an
accumulating conveyor from the inside set of conveyors (i.e., the BASICNODE), set the Input Buffer
capacity property to zero because it is not needed and there is no delay in off-loading to this conveyor.

Step 2: Add three BASICNODES to follow the SrvLabeling and name them Basic1, Basic2, and Basic3, in
that order. These are conveyor diverters that will send the sorted parts to their destinations. Add three
sinks, Snk1, Snk2, Snk3, and Snk0 to represent the destinations for parts of type 1, 2, 3, and 0
respectively. Connect up the objects with conveyors according to Table 12.1.

Table 12.2: Conveyor Properties

From/To Conveyor Type Conveyor Length Conveyor Speed
BasicNode4 to SrvLabeling Accumulating 10 meters 1.0 meters per second
SrvLabeling to Basic1 Non-Accumulating 10 meters 0.5 meters per second
Basic1 to Snk1 Accumulating 10 meters 0.1 meters per second
Basic1 to Basic2 Non-Accumulating 10 meters 0.5 meters per second
Basic2 to Snk2 Accumulating 10 meters 0.1 meters per second
Basic2 to Basic3 Non-Accumulating 10 meters 0.5 meters per second
Basic3 to Snk3 Accumulating 10 meters 0.1 meters per second
Basic3 to Snk0 Accumulating 10 meters 0.1 meters per second
Step 3: For the part’s. sequence table, replace the Input@SnkPartsLeave entries with
Input@SrvLabeling. Next add the sink node destinations associated with the part types. Figure 12.30
shows part of the table and you can insert the last placement at the end to maintain the sequence.

Figure 12.30: Changes to the Sequence Table

Next change the Output@SrvLabeling node Entity Destination to By Sequence. There is no need to add
intermediate nodes (the BASICNODES) since SIMIO can find the designations from the SrvLabeling.

Step 4: Select all the conveyors and add the conveyor path decorators as seen in Figure 12.31.

Materials Handling 203

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 12.31: Conveyor with Path Decorators
Step 5: Save and run the simulation for a while observing the animation which may need slowing down.

Question 21: Are the parts being conveyed as expected and do you see accumulation at the labeler?
________________________________________________________________________________

Step 6: Run the simulation for 40 hours.

Question 22: What is the average number of parts that accumulate on the conveyor in front of the labeler?
________________________________________________________________________________

Question 23: What is the Utilization of the labeler?

________________________________________________________________________________

Question 24: Why does the OUTPUTBUFFER of the labeler have content?
________________________________________________________________________________

Question 25: What is the average number on the conveyor between Basic2 and Basic3?
________________________________________________________________________________

Question 26: Do you have any concerns about the operation of the sorting conveyor system?
________________________________________________________________________________

Part 12.7: Commentary

· Material handling can be a substantial cost of production and thus is of great interest to many
companies. Vehicles and conveyors provide considerable flexibility in modeling material handling.

· The modeling of cranes has not been included in this chapter. These are a complex category of
materials handling. However, SIMIO has developed a special library for cranes that can represent a
wide variety of overhead and floor cranes. In particular, there is a 3D animation of the cranes to reflect
the fact that cranes operate in three dimensions. This package is highly recommended.

204 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Chapter 13
Management of Resources: Veterinary Clinic
The RESOURCE object provides a flexible means to model individual resource needs. Unlike the SERVER, the
RESOURCE does not provide a processing capability with its input and output buffers at a specific location.
Further because the RESOURCE object is not location restricted, as the units of its capacity can be used at
several places within a model. Because the resource is flexible, its capacity can be allocated (seized and
release) in a complex fashion. The management of a firm’s resources is one of main challenges of any
organization. A simulation of that management can greatly assist in determine how those resources can be
best used.

To illustrate part of the resource management capability, a Veterinary Clinic is modeled. The veterinary clinic
provides care for many different kinds of animals. The office wants to provide efficient care through the
employment of efficient resources. For simplicity we model primarily the activities of the veterinary staff.
The staff consists of four veterinarians named (first names) Billy, Jim, Maisy, and Ellie. This clinic service is
complicated by several factors. There are times when multiple types of resources, such as a Veterinary staff
member and an X-ray machine, are required before the patient serviced. There are certain preferences for
resources such as the fact the Jim is the only veterinarian who can service iguanas. The SERVER object cannot
handle these cases easily. For instance, you may need a nurse and doctor to perform a procedure and then need
the nurse for some after-procedure service.

Part 13.1: Utilizing the Fixed Resource Object

The RESOURCE object is a capacitated “Fixed object” (i.e., it does not visibly move in the network like the
TRANSFERNODE or SOURCE objects), and which can be seized and released. It does provide a generic object
that can be used to constrain entity flow. RESOURCES have reliability capabilities as well as the ability to
statically and dynamically rank and select the entities for processes (i.e., process all critical ill patients first) as
well it can choose to remain idle under some conditions. (i.e., refuse to be seized).

Step 1: Create a new fixed model (“Project Home” tab). Note, you can have multiple models in the same
project, but only one will run at a given time unless you use them “inside” of each other.

Step 2: Add a new SOURCE named SrcPatients, SERVER named SrvVetClinic, and a SINK named SnkExit.
Add a MODELENTITY named EntPatients. Connect the objects with paths that are logically 30 meters
long.173

· Patients arrive Exponentially with an interarrival time of 6.5 minutes.

· Each patient takes between 10 and 30 minutes to be seen by the staff with most patients taking 15
minutes to be served (i.e., use a Pert distribution to model the processing time).

Step 3: From the Standard Library section, add a new fixed RESOURCE174 named ResBilly.

· Since RESOURCES have reliability logic set up emergency phone calls for the doctors. Specify the
Uptime Between Failures to follow an exponential distribution with a mean of six hours and Time to
Repair expression equal to a Random.Pert(40,55,60) minutes as shown in Figure 13.1.

173Recall, if you select the output node of the SOURCE while holding down the Shift and Ctrl keys, SIMIO will allow you draw LINKS
without first selecting a LINK type from the Standard Library. This key sequence can be more efficient. Once the two PATHS have been
drawn, select them both and change their Drawn to Scale properties to “False” and the Logical Length properties to 30 meters.
174
The default symbol looks like a gray eye with a blue iris.
Management of Resources: Veterinary Clinic 205

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 13.1: Reliability Logic to Handle Emergency Call from the Veterinarian
· Like the servers, the resource symbol changes as its state changes. If you click on ResBilly and
examine the “active symbol”, there are seven resource states. The first five states are the same as
those for servers, but there are two additional states (i.e., “Failed Busy” and “Off Shift Busy”). The
symbols change colors, while the iris remains blue (i.e., the color is red when a resource fails).
Similar to the server states (see Chapter 8), the resource states have a name, color, and state number.
· Like the servers, the resource is idle (state 0) when all the capacity of the resource is idle, is busy
when any of the capacity is being used, blocked whenever the resource is presented from disengaging
from an entity, failed whenever the reliability causes a failure of the resource, and offshift when the
capacity of the resource is zero.
Step 4: Copy the ResBilly resource instance three times naming them ResEllie, ResJim, and ResMaisy
respectively to represent the other three staff members as seen in Figure 13.2.

Notice the symbol

colors for
resources

Figure 13.2: Model of the Veterinary Clinic Using Four Resources

Step 5: Change the capacity of the SERVER SrvVetClinic to be Infinity so the SERVER’S capacity does not
constrain the system. We want the new resource objects to be the constraining requirement.

Step 6: At this point, we can seize an individual RESOURCE but our model will require selecting among any
available staff member rather than a specific resource. Go to the “Definitions” tab of the model and create
a new LIST of OBJECTS named ListResources and add all four resources to the list as seen in Figure 13.3.
This list will be used by patients to select an available veterinary staff.

Figure 13.3: Creating a List of Resources to be Seized

206 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 7: The basic model has been set up with the fixed RESOURCES; however, the patients need to request
(seize) an available staff member to be serviced and then, once they have completed service, need to
release the staff member to work on the next patient. Patients will request a staff member when they arrive
to the Vet Clinic. Two Add-On Process Triggers for the SERVER will be used.

Step 8: The patient entity needs to seize one of the available resources before entering processing (i.e., the
“Before Processing”175 trigger). To invoke the “Before Processing” add-on process, expand the “Add-On
Process Triggers” in the properties window and double click on the label to invoke/create a process.176

Step 9: Since the patient wants to seize the resource; insert a Seize step by selecting from the list of
possible steps and dragging it between the Begin and End nodes which represent where the process
starts and ends as seen in Figure 13.4.

Figure 13.4: Using a Seize Step to Request a Veterinary Staff Member

· Under the “General” properties of the process step as seen in the right side of Figure 13.4, set the
Name to SeizeStaff. The resources to try to seize will be specified under the “Basic Logic”. Click on
the button in the property window in the Seizes property row which currently indicates that one
type of resource has been specified to be seized.

· The Seize – Repeating Property Editor (see Figure 13.5) will pop up and allow you to specify as many
resources that need to be seized before the entity can be processed. In this example, we will specify
one resource type and that the Object Type will be “FromList”177 can be selected from any in a list of
resources (i.e., ListResources). We do not want to select a resource that is currently failed so the
Selection Condition property should be !Candidate.Resource.Failure.Active.

Figure 13.5: Seize Property Editor

175
The Before Processing add-on process trigger is invoked after the entity has seized capacity of the server but before the entity is
physically transferred to the processing station while the Processing add-on trigger is invoked after the entity has been transferred and
directly before the delay for processing. The After Processing add-on process trigger is invoked directly after the processing has
finished.
176 Note, we can utilize the secondary resources to seize and release additional resources (i.e., Resource for Processing) using the same
list. However, utilizing processes gives you more flexibility and control.
177The Object Type for theseize can be “Specific”, “From List”, or “Parent Object”. The “Specific” option allows you to select an
individual RESOURCE (e.g., Jim) or any object that has capacity (e.g., Servers, Paths) or any of the Stations inside objects which contain
the capacity. The “Parent Object” will allow you to seize capacity of the object that is running the process which for the SERVER in this
process is the Processing STATION. The list option will specify any list that contains objects that can be seized.
Management of Resources: Veterinary Clinic 207

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 10: The patients are now able to Seize the clinic staff but once the patient has finished processing, the
entity (patients) will need to release the staff members (resource). There are two potential triggers (i.e.,
“After Processing” or “Exited”). Insert a new “After Processing” add-on process, by double clicking on
the label to invoke/create a process which will be executed once the process has completed before exiting.

· Insert a Release step into the process between the Begin and End nodes which will allow the
patient to release the resource seized, as seen in Figure 13.6.

Figure 13.6: Using a Release Step to Release a Veterinary Staff Member

· Repeat a similar process to the Seize step of naming the step as ReleaseStaff and specifying the
Releases property. Similar to the Seize Editor, the Releases → Repeating Property Editor will allow
one to release resources according to the release properties.

Figure 13.7: Release Property Editor for the Release Step

Step 11: Save and run the model for 40 hours and observe what happens.

Question 1: What is the overall average utilization for each of the RESOURCES and why does Billy have the
highest utilization?
________________________________________________________________________________

Question 2: What is the total time a patient has to wait for a staff person as and what is the average
number in queue for the input buffer of the clinic?
________________________________________________________________________________

Question 3: What is the total cycle time for the patients?

________________________________________________________________________________

Part 13.2: Different Resource Needs based on Different Patient Types

The previous model assumed all patients can be seen by all clinic staff members. However, suppose Billy and
Maisy have expertise with birds while Jim has expertise with iguanas, but if he is busy Ellie can provide the
service. Horses and cows are troublesome animals owing to size and require any two veterinarians. All other
animals (cats and dogs) can be seen by any one of the veterinarians.

Step 1: Create a new data table named TableAnimals that has a “String” column(property) for the animal
name, an “Integer” column for animal type, and a “Real” column for the percentage of time this type of
animal typically arrives to the clinic. The entry values are shown in Table 13.1. For example, 25 percent
of the time large animals representing cows and horses will arrive at the clinic.

208 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Table 13.1: Table with Animal Information and Types

Step 2: Create two new OBJECT lists named ListBird and ListIguana in a similar manner as
ListResources in Figure 13.3 that represent the staff that can service these specialty animals as seen in
Figure 13.8.

Figure 13.8: Object Lists for the ListBird and ListIguana

Step 3: Add three additional symbols to the patient entity to distinguish between the various animal types.
One can use four different colors but you can also use interesting pictures of these creatures by
downloading symbols from Trimble 3D Warehouse as seen in Figure 13.9 which shows the four symbols
horizontal (instead of vertical). Note the order of the symbols should correspond to the type column in the
Table 13.1.

Figure 13.9: Trimble 3-D Warehouse Downloaded Dog, Cow, Bird, and Iguana Symbols
Step 4: The goal will be to randomly assign the animal types and pictures (color) based on the historic
percentages. The assignment must be done after the entity has been created using the “Created” add-on
process trigger in the SOURCE. Insert a new “Created” Add-On Process Trigger178 which will be executed
once the entity has been created by the SrcPatients.

· One of the features of SIMIO is the ability to associate one or more tables and/or rows of a table
with an entity. Our newly created patient needs to be associated with a specific row of the
TableAnimals data table and then assigned the correct picture symbol based on the “Animal
Type” column. Insert a Set Row and an Assign process steps into the process as in Figure
13.10.179

Figure 13.10: Determining Patient Type by Setting a Row of Datatable to Entity

178
To invoke the “Created” add-on process, expand the “Add-On Process Triggers” of the SOURCE in the properties window and
double click on the label to invoke/create a process.
179
In this example, the Table Reference Assignments (i.e., On Created Entity) could be used to assign a table. Without knowledge of
the object you cannot be sure that the picture assignment via State Assignments happens before or after the table assignment but with
the processes approach you can control the order.
Management of Resources: Veterinary Clinic 209

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 · Specify a specific row of the associated table randomly using the RandomRow method based on
the “PercentType” column as seen in Figure 13.11.180 Also make sure the Object Type under the
Advanced Options is set to the “AssociatedObject”.181

Figure 13.11: Specifying the Properties of the Set Row Process Step
· Once an animal type has been determined, Figure 13.12 shows the properties of the Assign process
step used to assign the associated “Type” column from the table to the picture symbol state of the
MODELENTITY.182

Figure 13.12: Changing the Picture of Entity to Reflect a Column in an Assigned Datatable
Step 5: Save and run the model to make sure the different animal types arrive the way they were expected.
One may need to adjust the animated processing queues to allow the large animals to be visible.

Step 6: Currently, the patients will still seize any resource regardless of animal type, but only one staff
member is seized. Therefore, we need to modify the Add-On Process trigger for seizing and releasing
resources to incorporate the resource requirements (See flow chart in Figure 13.13),

· If the type is cat/dog or horse/cow then we can chose any resource, but for horse/cow patients
two resources are needed.
· If the type is bird we need to choose a resource from the OBJECT LIST ListBird
· Finally if the type is iguana we will need to choose from the OBJECT LIST ListIguana

180Any numeric column can be used to specify a random row of table. The numbers do not need to add up to 1 or 100. SIMIO will take
the sum of the column and then generate a random row based on the normalized percentages.
181
The “ASSOCIATED OBJECT” is the object the current token is represents. In this case, it is the entity or patient. The other options are
“TOKEN” for the current token executing the process, the “PARENTOBJECT” which would be the SOURCE since it owns the Add-on Process
Trigger, and “SpecificObjectOrElement” which allows you to set the table on a particular object or element.
182
Notice, the animal types were specified on purpose to be zero to three to represent the symbol numbering for the pictures. Also, the
green arrow lets the user know this refers to a referenced property.
210 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 True Seize 1 or 2
If Patient is a
Begin Resources from End
Dog or Cow
ListResource

False

True Seize 1
If Patient = Bird Resources from
ListBird

False

Seize 1
Resources from
ListIguana

Figure 13.13: Seizing the Appropriate Veterinary Staff Based on Patient Type
· To perform the logic we need to make some decisions based on the animal type as seen in the above
flow chart. The logic is performed by a SIMIO Decide process step (see Figure 13.14) which is
basically an If/Else statement that has a True and False branch. In Figure 13.13, the Decide step’s
two major properties of interest are the Decide Type and the decision Expression.

Figure 13.14: SIMIO Decide Process Step

Figure 13.15 shows the modified “Before Processing” add-on process trigger that will seize the
appropriate staff members based on animal type.183

Figure 13.15: Seizing the Appropriate Resource Based on Animal Type

· Insert a Decide step before the current Seize step which will determine if the patient is a dog/cat or
cow/horse. Based on which is true, the patient will seize from the resource object LIST ListResource
as before. Specify the Decide Type to be condition based with an expression that checks to see if the
animal type is a 0 or 1 (see Figure 13.16). 184

183
The Processing add-on process trigger occurs after the entities are transferred into the processing station so the animals will appear
in the processing station even though they have not seized a vet. Therefore, use the Before Processing add-on process trigger to
eliminate the animation issue.
184The expression (TableAnimals.Type < 2) determines if the type is a zero or one could have been specified
(TableAnimals.Type == 0 || TableAnimals.Type ==1) where || equals the Or logical operator.
Management of Resources: Veterinary Clinic 211

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 13.16: Specifying the CondtionBased Expression of a Decide Step
· The second Decide step will check to see if it is a bird.( Hint, just copy the first Decide block and
paste it onto the False branch and change the expression to (TableAnimals.Type==2)).
· Copy the Seize step to both the True and False branches of the second Decide step. You will need
to change which resource list the entity will seize from (i.e., ListBird and ListIguana respectively).
· The first Seize step is already setup to seize from the all the resources (ListResources). However,
by default, it only seizes one object. But two resources are needed for the large animals. There are
several ways to model the logic. Since the animal type of cats/dogs is zero and cows/horses is one, the
type can be added to one to give a result of one or two. In the Number of Objects185 property to seize,
the following Figure 13.17 illustrates the expression to seize one or two objects based on the animal
type.

Number of
Resource Objects

Figure 13.17: Specifying an Expression to Determine the Number of RESOURCES to Seize

Step 7: Save and Run your model. You may need to experiment with the percentages to make sure the
animals are being serviced by the correct staff members. Figure 13.18 shows a cow being serviced by both
ResBilly and ResJim.

Figure 13.18: Using Two Resources to Service a Large Animal.

Question 4: Did the staff get released and why?
________________________________________________________________________________

185 The Number of Objects property specifies the unique number of objects to select out of a list while the Units Per Object specifies
the number of each object to seize. For example, if your list had doctors and nurses and you set the Number of Objects to “1” and Units
Per Object “2”, then it would seize either a doctor or nurse object but it would require two doctors or two nurses. Resources can be
given capacity of more than one (e.g., you could have ten nurses but you do not want to have ten different nurses (i.e., Billy, Ellie,
etc.).
212 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 8: The only thing left to do is to release the veterinary staff once they have finished processing the
patient. It is similar to the “Before Processing” trigger but uses Release instead of Seize steps. It is
important that you release the correct number of objects back to the ListResources pool.

Figure 13.19: Releasing the Correct Vet to Appropriate Resource Set

Step 9: Save and Run your model and observe what happens (see Figure 13.20).

Animals being
served

Animals
waiting

Figure 13.20: Figuring Showing Animals being Serviced and Waiting

Question 5: What is the overall average utilization of each of the RESOURCES?
________________________________________________________________________________

Question 6: What is the total time a patient has to wait for a staff person as well as the average number in
queue for the input buffer of the clinic?
________________________________________________________________________________

Question 7: What is the total cycle time for the patient?

________________________________________________________________________________

Part 13.3: Resource Decision Making

One of the most prominent features of SIMIO is that you can model very complex logic (e.g., process steps) or
expressions, and supplement it by associating information via data tables on entities. In an earlier chapter, the
Management of Resources: Veterinary Clinic 213

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 processing time and routing information for different part types was linked directly to the part itself. When
parts arrived at the SERVERS, the server would ask the entity how long to process it. Many simulation
languages, including SIMIO, have the ability to specify attributes (i.e., entity state variables in SIMIO) which
can be assigned the actual processing time value when the entity is created. However, SIMIO information in a
data table can be dynamically changed for any object to reference to it. In the current example, there are four
different animal types. What if the number of types increases to ten with eight or nine different lists to seize
the correct vet? The decision logic would get very complicated and long. Therefore, we have the ability to
make the objects more intelligent (i.e., they know the list and the number required for service).

Step 1: Create or open the model from the previous section and save it as a new name.

Step 2: Recall that only the animal Type column was utilized previously to make decisions. Now modify
the data table TableAnimals by inserting a new “Integer” column from the Standard Property dropdown.
Name this column NumberStaff to represent the number of vets that are required by the animal type.
Next add an OBJECT LIST column from the Object Reference dropdown named ListStaff which will
represent the list the resources that can be selected. Set the rows values according the ones in Figure
13.21. Now the patients know which group of people can service them as well as the number that are
needed.

Figure 13.21: Including the Number of Staff Needed in the TableAnimals

Step 3: In the “Processes” tab remove all the steps except one Release and Seize step in the “Before
Processing” and “After Processing” add-on triggers respectively as seen in Figure 13.22.186

Figure 13.22: Improved Process Logic for Seizing and Releasing Vet Staff Members
Step 4: Modify the properties in both the Seize and Release steps to utilize the referenced/associated
table properties. On the Object List Name and Number of Objects, right click and set the Referenced
Property to the correct column of the table respectively. Now when a patient arrives to seize a staff
member, it will dynamically seize the appropriate number from the correct group of resources.

186
Modeling the logic in this fashion has removed the two decision steps which will speed up the simulation model as well as simplify
the logic when, for example, more animal types are added. Now that information is contained in the data table, eliminating the need to
add additional process logic.
214 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 13.23: Seize and Release Properties using Referenced Properties
Step 5: Save and run your model and compare the results to previous model’s results.

Part 13.4: Adding an Additional Process

Sometimes, the patients need to have an X-ray taken which is located in a different room. For example, 25%
of cats and dogs typically need X-rays while only 15%, 0% and 2% of cows and horses, birds, and iguanas
respectively do. The patients will leave the patient care rooms and be moved to the X-ray machine room. The
veterinarians will accompany the patients to perform the X-ray. Afterwards the veterinarians will be released
to serve other patients.

Step 1: Add an additional SERVER named SrvXrayMachine as seen in Figure 13.24.

· Set the capacity of the SERVER to one since there is only one machine
· The processing time to perform and read an X-ray follows a Pert distribution with a minimum, most
likely, and maximum values of 20, 25, and 35 minutes respectively.
· Connect the output of the X-ray machine to SnkExit via a 30 meter logical PATH.
· Connect a 10 meter logical PATH from the SrvVetClinic output to the input of the SrvXrayMachine.

Step 2: Since each animal type has a different chance of needing an X-ray, add an additional Real column
named XrayLabPercentage to the TableAnimals as seen in Figure 13.25.

Figure 13.24: Adding an Additional Process that uses Resources Already Seized

Figure 13.25: Animal Percentages of Needing an X-ray

Management of Resources: Veterinary Clinic 215

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 3: If the patient needs an X-ray, the staff member(s) should not be released immediately after the
patient care service. The entity should move to the X-ray station SERVER where they will be processed in
the order they arrive. Once they have completed the X-ray process, the staff members are released back to
the pool of available veterinarians and the patients will exit. The first step will determine if the patient
should release the staff members or not (i.e., a Decide process step will be needed). In the “After
Processing” add-on process trigger of the SrvVetClinic insert a Decide process step. Move the
Release step to the “False” branch (see Figure 13.26) so that if an X-ray machine is needed, then the
staff is not released.

Figure 13.26: Modified Processed Process Trigger

· The Decide step is shown in Figure 13.27.

Figure 13.27: The Decide Step

Step 4: If the patient needs an X-ray, they release the staff members after the X-ray has finished processing
them. Insert a new process for the After Processing add-on process trigger for the SrvXrayMachine.
Copy and paste the Release step block from the SrvVetClinic_AfterProcessing process to the new
process.

Step 5: Now the patients who need X-rays need to be routed to the new SERVER. By default, the patients
will randomly go 50% of the time to the sink or the input at the X-ray machine owing to the equal link
weights of one.187 There are numerous functions (i.e., f(x) designation in the expression editor)
associated with the various objects. Table 13.2 gives a description of the SeizedResources family of
functions. The base function of SeizedResources which gets access to the objects that any intelligent
object has seized will invoke the underlying function SeizedResources.NumberItems.

187
Note, we could have specified the link weights to be 1-TableAnimals.XrayLabPercentage for the link to the sink and
TableAnimals.XrayLabPercentage for the link to the X-ray machine and the patients would have been routed correctly.
However, this logic would only work if the patients did not need the same staff members to accompany them to the X-ray machine
(i.e., they could be released before exiting the SrvVetClinic).
216 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Table 13.2: Description of all the SeizedResources Functions
Function Description
This function will return the current number of objects (i.e.,
SeizedResources.NumberItems RESOURCES, SERVER and LINK capacities, etc.) seized and owned
by the object using the function.
All object instances in a SIMIO model have a unique numeric
identifier (ID) (e.g., NODES, SERVERS, RESOURCES, etc.) This
SeizedResources.FirstItem
function will return the reference of the first object in the list of
objects currently seized and owned by this object.
This function will return the reference of the last object in the list
SeizedResources.LastItem
of objects currently seized and owned by this object.
This function will return the reference of the object in the list of
SeizedResources.ItemAtIndex
objects currently seized and owned by this object at the specified
(index)
index.
This function will return the index of a specified resource in the
SeizedResources.IndexOfItem
list of resources currently seized and owned by this object. It will
(resource)
return zero if the resource has not been seized.
This function will allow you to determine the number of units of
SeizedResources. capacity for a specified resource that are currently owned by
CapacityOwnedOf(resource) this object (i.e., it would return two if they had seized two
hammers).
This function returns True(1) if the list of resources currently
SeizedResources.
seized and owned by this object includes the specified resource.
Contains(resource)
Otherwise the value False(0) is returned.
· Specify the link weight of the path leaving the SrvVetClinic and going to the exit (SnkExit) to be
ModelEntity.SeizedResources.NumberItems == 0 which means the patient has
released the staff members and can exit the system.
· Specify the link weight of the path leaving the SrvVetClinic and going to the X-ray machine
(SrvXrayMachine) to be ModelEntity.SeizedResources.NumberItems > 0 which
means the patient has still seized staff members because it needs to have an X-ray done.
· Note the link weights will evaluate to one or zero thus forcing the patients to take the path only if the
link weight was a one.

Step 6: Save and run the model and observe what happens.

Part 13.5: Changing Processing Based on Animal Type and Vet Servicing

At this point, all of the animal types take the same amount of time to be seen by one of the vets (i.e.,
processing time is sampled from the same distribution). However, depending on which animal type as well as
which vet is seeing the animal the time to be seen can vary. This is a common situation where the time is
dependent on the resource that was seized (i.e., doctor or nurse, etc.). Recall that relational data tables were
used previously to specify a different processing time for each of the part types for each of the operations in
the sequence. This method works well for distinguishing processing times among different part types (i.e.,
animal types). Therefore, a child table containing the processing times will be created to allow the processing
time to become also dependent on the resource being employed.

Step 1: First, select the Type column in the TableAnimals and make it the primary key of the table as seen
in Figure 13.28. This table will become the parent table from which the child table will inherit the values.

Management of Resources: Veterinary Clinic 217

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 13.28: Specifying the TableAnimals Type Column to be a Primary Key.
Step 2: Next add a new data table named TableProcessingTimes via the Data→Tables→Add Data Table
button as seen in Figure 13.31.

· Add a new Standard Expression Property column named ProcessingTime as seen in Figure 13.29
making sure to set the properties Unit Type to “Time” and Default Units to “Minutes.”188

Figure 13.29: Specifying the Processing Time Column

· Next, add a new Foreign Key column named Type with the Table Key property set to
TableAnimals.Type as seen in Figure 13.30.

Figure 13.30: Specifying the Foreign Key to inherit Column from Parent Table
· At first we will utilize constant times for each of the animal services. Set the time to 30, 60, 20, and 40
minutes for dogs, cows, birds and iguanas respectively to correspond to Figure 13.31.

Figure 13.31: Processing Time Table for the Different Animal Types
Step 3: Next we need to utilize the new processing time information into the model. Unlike prior cases
where we were using the next sequence of the TRANSFERNODE to set the particular row in the child table,

188
It has to be an Expression property to allow distributions to be specified.
218 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 we will need to specify it directly (i.e., set the Processing Time property for row one by
TableProcessingTimes[1].ProcessingTime of SrvVetClinic). Without the relational table,
the expression would set it to the first entry in the whole table always. Since this is a related table, it will
set it to the first entry of the inherited rows which currently only has one row as seen in Figure 13.33
which shows for child table for an entity which has been assigned a bird. 189

Figure 13.32: Setting the Processing Time to be the First Row of the Child Table

Figure 13.33: Showing the Child Table Associated with the Bird Row
Step 4: At this point we could run the model but it may be difficult to determine if the processing time is
being assigned correctly.190 Therefore, we will set the priority191 state variable of each entity to reflect
their processing time. In the SrcVetClinic_Processing process, insert an Assign step after the Seize
step to set the patients PRIORITY state variable as seen in Figure 13.34.192

Figure 13.34: Specifying the Processing Time for Display Purposes

Step 5: In order to see the set priority value, we will add a status label to the animal that displays the
processing time for each animal.193 Select the MODELENTITY EntPatients and then insert a new
Animation→Status Label which associated/attaches the label to the entity. Specify the expression of the
label to be the entity’s Priority state variable.

Status Label Attached to ModelEntity

Figure 13.35: Inserting a Status Label Attached to the Entity

189
Notice, an error would occur if we just specify TableProcessingTimes.ProcessingTime as was done in a previous
chapter without first setting the row via a Set Row step or a By Sequence from the TRANSFERNODE.
190 We could turn on the trace and watch what the processing time delay was when the animal went into service.
191
Recall that “priority” is a SIMIO-defined state variable for entities.
192
We are multiplying by 60 because SIMIO has converted all times to hours internally.
193SIMIO now provides a Dynamic Label Text property under the Animation section which can be used to a similar thing but the label
appears below the MODELENTITY.
Management of Resources: Veterinary Clinic 219

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 8: Why do we have to multiply the processing time by 60?
________________________________________________________________________________

Step 6: Save and run the model observing the status label times.

Question 9: Are the correct times being assigned to the different animals?
________________________________________________________________________________

Step 7: Now the model can handle the situation where the processing time is based on the animal type. In
many situations the processing time may also be dependent on the resource (i.e., vet) you have seized. The
notion of position within the list of resources will be used to determine the processing time. First, select
the MODELENTITY in the [Navigation] panel and add a new DISCRETE INTEGER STATE variable from the
Definitions→State section named EStaWhichRes. This state variable will be assigned the position of the
resource that has been seized in the particular resource list.

For example, if a current patient who is a dog has seized Ellie from the ListResource list the patient’s
EStaWhichRes variable would be assigned a value of “2” to correspond that ResEllie is in the second
position of this list as seen in Figure 13.36.

Position ListResource
1 ResBilly
2 ResEllie ModelEntity.EStaWhichRes set to 2
3 ResJim
4 ResMaisy
Figure 13.36: Illustrating how Resource Position will be used
Step 8: Currently the data table TableProcessingTimes contains only time for each animal type. Modify
the table to contain a different processing time for each vet (i.e., resource) and each animal type as seen in
Figure 13.37.194 Since four different resources are capable of processing dogs, there are four different
processing times (i.e., 31 minutes for Billy, 32 minutes for Ellie, 33 minutes for Jim, and 34 minutes for
Maisy) for animal type “0” (i.e., cats and dogs). For birds and iguanas only two vets can see each of those
types of patients, therefore only two different processing times are specified for types “2” and “3”. As
Figure 13.37 illustrates, Jim will process Iguanas in 42 minutes while it takes Ellie 43 minutes.

194
One of the easiest ways to create the table is to copy it from an Excel worksheet where the table can be created easily. Tables can be
copied to and from Excel. To copy from Excel, first copy the two dimensional table in Excel and then highlight all the current values in
the SIMIO data table before pasting the new values which will overwrite any existing one and then create the any new rows as needed.
220 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Billy
Cats & Ellie
Dogs Jim
Maisy

Cows &
Horses

Birds
Jim
Iguanas Ellie

Figure 13.37: Add more Processing Times for Each Resource

Figure 13.38 shows the child related table for animal type zero (i.e., cats and dogs) when expanding out the
type “0” from the parent table TableAnimals. As you can see, only four processing times are related to type
zero. Now going back to the example in Figure 13.36, Ellie would corresponds to the second position in the
list which translates to row two of the related table (i.e., 32 minutes).

TableProcessingTimes[ModelEntity.EStaWhichRes].
ProcessingTime

Ellie

Figure 13.38: Showing the Child Sub Table for Cats and Dogs Animal Type
Step 9: The only step left is to appropriately assign the EStaWhichRes variable based on the resource that
was seized. To accomplish this task, we will employ a Search process step where Table 13.3 explains all
of the properties associated with the step.195 The Search process step can search a collection of items
(see Table 13.4) and return TOKENS associated with the items in the collection.

195 Recall the TOKEN executing the process steps is typically associated with an entity. Therefore any steps (i.e., Transferring, Setting

Nodes, etc.) are only applied to the MODELENTITY associated with the TOKEN. Often we may want to affect other objects (resources
seized, etc.) besides the current entity, the search step can be used to create new TOKENS associated with other objects. In some
process steps, the action can be applied to the parent object of the TOKEN’S associated object (e.g., the SERVER the logic resides).
Management of Resources: Veterinary Clinic 221

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Table 13.3: The Properties of the Search Process Step
Property Description
Collection Type Type of the collection to search (See Table 13.4 for more information).
The type of search to perform (Forward, Backward, MinimizeReturnValue, and
Search Type
MaximizeReturnValue)196
Optional match condition that can be used to filter the items in the collection before
Match Condition
the search begins. Don’t forget to use the keyword CANDIDATE in the condition.
An expression that is evaluated for each item found by the search. The total sum for
Search Expression this expression is returned in the ReturnValue of the original TOKEN. One could
save the particular Resource and Seize the exact same one later.
One based index to start searching. By default it is one for Forward searches while
Starting Index
the number in the collection for Backward searches.
Ending Index One based index to end searching to assist in narrowing the search within collection.
An expression that specifies the maximum number of items that can be found and
Limit
returned (e.g., Infinity or one)
Save Index Found Specify an optional discrete state variable to save the last item found
Save Number Found Specify an optional discrete state variable to save the number of items found
Table 13.4: Different Types of Collections that Can Be Searched
Collection Type Description
EntityPopulation Search a particular entity population (i.e., MODELENTITY, WORKER, or VEHICLES)
QueueState Search the objects currently in one of the queues
TableRows Search the rows in a Data table
ObjectList Search the objects in a list (e.g., ResourceList, etc.)
NodeList Search all nodes in a particular list.
TransporterList Search all transporters in a particular list.
SeizedResources Search the objects currently seized by the Parent Object, Associated Object or a
Specific Object
ResourceOwners Search the objects that currently own (i.e., seized) a Parent Object, Associated Object
or a Specific Object. Search all animals who have seized the SrvVetClinic capacity.
NetworkLinks Search all links of a particular network.
NodeInboundLinks Search all the inbound links of a particular node.
NodeOutboundLinks Search all the out bound links of a particular node.
ObjectInstance Search all objects of a particular class (e.g., all entities that are currently created).
Insert a Search Step between the Seize and the Assign (see Figure 13.39). The “Original” branch is
associated with the original TOKEN (i.e., customer) while the “Found” branch will be invoked for a TOKEN
associated with each item found.197 The collection type will be an “object list” with the list name based on the
animal type (TableAnimals.ListStaff). We only want to match the last resource seized by the patient with a
resource from the resource list.198 If it finds a match, it will save the index (i.e., row position) in the
MODELENTITY’s StaWhichRes state variable.

196
MinimizeReturnValue and MaximizeReturnValue search values will search the collection and return up to the limit of items
specified that minimizes or maximizes the Return Value expression property.
197 The original token will not move forward until all the searching is done.
198
Note the use of the CANDIDATE keyword to represent each member of the collection being searched (i.e., like a wildcard *). Also, in
the case of a cow or horse which requires two vets, the second vet seized will be lower in the list and our assumption would reflect less
seniority and therefore slower.
222 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 13.39: Adding a Search Step to Find the Index Position of the Seized Resource in the List
Step 10: Now that the EStaWhichRes has been set, you need to modify the Processing Time property of the
SrvVetClinic as well as the Assign step to utilize the following expression (i.e., change the row index of
“1” to ModelEntity.EStaWhichRes).
TableProcessingTimes[ModelEntity.EStaWhichRes].ProcessingTime
Step 11: Save and run the model observing the status label times to see if different processing times are
being assigned based on animal type (i.e., in the 30, 60, 20, and 40 minute ranges) as well as which vet
(i.e., resource is serving the patient).

Part 13.6: Changing the Resource Allocation Selection

In the original description, Jim was the preferred staff member to see Iguanas while Ellie could see them if he
was busy. The clinic is considering a new policy that when Jim becomes available, he will check to see if an
Iguana is waiting and if so, he will service these patients first before seeing the next patient. A “first-in and
first-out” ranking scheme is currently being used by the SrvVetClinic to service the patients.
RESOURCES like SERVERS have a static ranking rule199 which will order the requests waiting to be allocated
capacity from this object. The SERVER and RESOURCE will process them in the order the entities are sorted.200
Step 1: Change the static ranking rule of the RESOURCE ResJim to be the “largest value first”. The ranking
expression should be the referenced property animal type since iguanas will always be serviced first if
they have requested service.201 The static ranking rule works because the type of the animal does not
change, while the patient is waiting for a vet (see Figure 13.40).

199
The static ranking rules are First-in First-out (Default), Last-in Last-Out, Smallest Value, and Largest Value, where an expression
provides the value for the last two rules (e.g., ModelEntity.Priority)
200
Also there is a dynamic selection rule, which will be considered in a later chapter.
201
If ResJim is busy processing a patient and an iguana is at the front of the server’s queue then ResEllie may still see them, which is
acceptable. The goal is to maximize ResJim’s chances of seeing Iguana patients but not to eliminate her’s completely. If only ResJim
could see iguanas then the ListIguana could be changed to only have Jim as a resource.
Management of Resources: Veterinary Clinic 223

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 13.40: The Largest Value Static Ranking Rule
Step 2: To visualize the ranking of ResJim, select ResJim and insert an animated queue (i.e.,
ALLOCATIONQUEUE) from the Draw Queue section of the Symbols tab. The allocation queue is the queue
that contains the requests for allocation of capacity of an object (i.e., RESOURCE, SERVER, WORKER,
VEHICLE, COMBINER, etc.). Make the queue long enough to accommodate the size of the animals as seen in
Figure 13.41 which is drawn from right to left.

Jim’s AllocationQueue

Figure 13.41: Allocation Queue of ResJim

Step 3: Save and run the model and observe what happens. The percentage of time Iguanas arrive may
need to be altered to assist in verifying the model works the way it is intended.

Question 10: Are there any issues with ranking the queues in this manner?
________________________________________________________________________________

Question 11: Did you observe cows jumping ahead of dogs in the queue?
________________________________________________________________________________

Step 4: Jim should service the Iguanas first and then service the remaining two animals in order of their
arrival. One way to accomplish this ranking is to modify the Ranking Expression property to the value in
Figure 13.42. The logical expression will make all iguana patients have a value of one; while all other
patient types would be zero and therefore preserve the arrival order to break the ties among remaining
animals (arrival order is the tie-breaking for static ranking rule).

Figure 13.42: Modifying the Ranking Rule to not Service Cows before Dogs
Step 5: Save and run the model and observe what happens now.

Part 13.7: Commentary

· A very common modeling technique is to add additional state variables to the general entity to hold
values or states. In the second part, an EStaXray discrete state variable could have been added to the
entity. In the True branch of the “After Processing” add-on trigger as seen in Figure 13.43, an Assign
step could have been added to assign the value of one to the new EStaXray state variable to show that an
X-ray is needed.

224 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 13.43: Using a State Variable to Assign Whether an X-ray is needed or Not
· The link weights could have been specified as (1-ModelEntity.EStaXray) and
(ModelEntity.EStaXray) for the links to the exit SINK and the X-ray SERVER machine, respectively.
Again the link weights will take on zero or one and will never be the same. If the patient does need an X-
ray, then (1-ModelEntity.EStaXray) would evaluate to one while (ModelEntity.EStaXray)
would evaluate to zero.
· To some extent, one could have used the Secondary Resources to seize and release the vets. One
could of easily used Before Processing under Other Resource Seizes and After Processing under Other
Resource Releases with the same input as Figure 13.23. However, the ability to release or not release if
the animal needed a vet would require a little more modeling if you were trying to avoid processes all
together. It is easy not to release the Vets at the clinic and then only release them under the After
Processing under Other Resource Releases of the SrvXrayMachine if the animal needed an X-ray.
However, there is no mechanism to release the vets for those who do not. You would have to add an
additional SERVER between the SrvVetClinic and the SnkExit that would release them. As has been seen
in other chapters, one has more control and flexibility in using processes.

Management of Resources: Veterinary Clinic 225

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Chapter 14
A Mobile Resource: The Worker
The previous chapter illustrated the flexibility of RESOURCES and RESOURCE LISTS to model very complicated
relationships. However, these types of resources were fixed, in the sense that they do not move dynamically
throughout the system model. When the patients need an X-ray, the resources were not released and the
patient moved to the X-ray machine room, but the animation did not show the resources moving along with
the patient. Also, in the DMV problem, resources were used in multiple locations but the travel time between
the two locations was not considered. SIMIO has the ability to have multiple types of the objects (i.e., entities,
workers, and vehicles) that can act as resources (i.e., objects that can be seized and released) and move
throughout the system, similar to an entity.

Part 14.1: Routing Patients

Again, we return to the veterinary clinic. However this time the clinic will behave more consistently with a
doctor’s office where patients arrive to the clinic, check in at the front desk, and then wait in the waiting room
for an available room. The patients will move to one of four examination rooms once one becomes available.
Then one veterinarian will start from their office and will dynamically move to the rooms as needed and
service the patients as requested. The veterinarian will be modeled as a movable resource that needs to be
seized and released.

Step 1: Create a new model that contains one SOURCE (SrcPatients), a SINK (SnkExit), and five SERVERS
(SrvVetWaiting, SrvRoom1, SrvRoom2, SrvRoom3, and SrvRoom4).

Step 2: Insert a new MODELENTITY named EntPatients into the model. The patients travel at a constant rate
of 4.5 km per hour or 4500 meters per hour.

Step 3: Insert the rooms into a two hall configuration with the waiting room serving the two hallways as
seen in Figure 14.1.202

· The capacities of all the servers should be one.

· The check-in process at the SrvVetWaiting uniformly takes between one and four minutes before
they are placed into the waiting room.
· Each patient takes between 10 and 20 minutes to be seen with most of them taking 12 minutes.
· It has been observed that patients arrive exponentially with an interarrival time of 6.5 minutes.

Step 4: Insert a wall in the middle of the rooms to add a more visual cue for the hallway separation as seen
in Figure 14.1. Make sure to specify a height of 3.5 meters meters203 as well as give the wall a texture. You
will notice we inserted a different line (i.e.., window texture) around the waiting room plus added a 3-D
door as the entrance.

202
To be more efficient, create one room named SrvRoom with all the correct properties and then copy it three times changing the
name of the original one to SrvRoom4 accordingly.
203
After you have selected the line or rectangle, the Object Height and Line Width text edit area can be found under the “Object”
section of the “Drawing” Tab.
226 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Separate Paths to
Exit Patients Connector

Four Individual Paths to

Rooms from Waiting Room

Figure 14.1: New Veterinary Clinic Model for Moveable Resources

Step 5: Connect the output of the SrvVetWaiting to each of the inputs to the four rooms using four PATHS
named PathRoom1, PathRoom2, PathRoom3, and PathRoom4 respectively.

· The distance between waiting room and rooms one and three is 15 meters and to the other rooms it is
30 meters.
· Select all four PATHS and set the traveler capacity to one and the speed to 4.5 kilometers per hour.
Step 6: Next, provide a means for the patients to leave the examining rooms and exit the clinic using paths
and the same distances from the waiting room to the rooms for the rooms to the exit. TRANSFERNODES can
be used to assist in directing the patients out as seen in Figure 14.2.204 We can use common paths (i.e.,
paths between room one and room two) for the output as we will need to constrain the paths to only have
one person traveling on the path as we will the input paths.

Step 7: Connect the SOURCE (SrcPatients) via a CONNECTOR to the SrvVetWaiting and use a CONNECTOR to
the SnkExit from the TRANSFERNODE

Step 8: Modify the OUTPUTBUFFER.CONTENTS queue of the SrvVetWaiting server to have an Oriented
Point alignment with several locations with the Keep in Place option checked which will improve the
animation.

Step 9: As before, patients should select a room that is currently available. Create a new node list via the
Definitions tab named ListRooms which will allow the patients to choose a room that is available. Figure
14.2 shows the list, noting the order of preference is one, three, two and then four (i.e., fill up the closest
rooms first if more than one room is available).

204 CONNECTORS are used to connect the outputs of the rooms to the TRANSFERNODES while paths with appropriate logical lengths are

used to connect the nodes.

A Mobile Resource: The Worker 227

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 14.2: Creating a Node List of Examination Rooms
Step 10: For the output TRANSFERNODE of the waiting room (Output@SrvVetWaiting), specify the Entity
Destination Type to select a node “From a list” using the ListRooms as the Node List Name property.
The goal is to route patients to an available room. Therefore, specify the Selection Goal as the “Smallest
Value”. In previous chapters we used an expression to determine the total number of entities routing to the
input node, waiting in the input buffer, plus those currently being processed by the server.

Step 11: However, a series of AssociatedStation205 functions (see Table 14.1) can be used as part of
the dynamic routing selection for all fixed objects that have external input nodes (e.g., SERVERS). These
external input nodes are associated with an internal STATION (e.g., Input@SrvRoom1 is associated with
the input buffer station of SrvRoom1).
Table 14.1: Explanation of the AssociatedStation Functions
Function Description
This function returns the current capacity of the location inside
AssociatedStation.Capacity the associated object (e.g., the capacity of the input buffer for
the server). 206
AssociatedStation.Capacity. This function returns the current available unused capacity of
Remaining the input buffer of the server.
This function will return the current number of entities in the
AssociatedStation.Contents location (e.g., number in the input buffer of a server
(InputBuffer.Contents).207
AssociatedStation.EntryQueue. This function will return the current number of entities waiting
NumberWaiting to enter the location (InputBuffer.EntryQueue).
This function will return the “load” which is the sum of
entities on their way to the node, waiting to enter the object,
AssociatedStationLoad
and now in the object (i.e., input buffer and processing
stations).
This function returns how much a location is overloaded
AssociatedStationOverload which equals the “Load” minus “Capacity” which can be both
positive (overloaded) or negative (under loaded).
Use the expression (Candidate.Node.AssociatedStationload) which returns the total number of
entities heading to the room, waiting to enter the room, waiting at the room plus the current number of entities
being processed by the room as the Selection Expression property as seen in Figure 14.3.

205
When you access the AssociatedStation functions for an external node, you are given access to the actual station and all of
its properties, states and functions.
206The capacity of the SERVER (i.e., the Processing STATION) is not included unless the InputBuffer capacity is zero then all the
functions report values associated with the Processing STATION instead of the InputBuffer STATION. Note that an infinite capacity is
represented by the maximum integer value of 2147483647.
207
The AssociatedStation.Contents for the input buffer does not include the entities that are currently being processed by
the SERVER or those in the OutputBuffer station. Also, the AssociatedStationLoad does not take into account entities
waiting in the OutputBuffer STATION.
228 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 14.3: Specifying the Selection Expression to Choose the Room with Smallest Number of Patients
Step 12: Save and run the model to observe what happens.

Question 1: Does the model behave the way it was intended?

________________________________________________________________________________

Step 13: Patients seem to wait at the examining rooms after being checked-in instead of at the waiting room,
which was not what was intended. Therefore, we need to force the patients to wait in the waiting room
until a room is available in a similar fashion as was done in the Chapter X. Select each of the rooms and
change the Input Buffer208 capacity to “0”. Select the output node of the SrvVetWaiting
(Output@SrvVetWaiting) and make sure the Blocked Destination Rule is set to “Select Available Only.”

Step 14: Run model to observe what happens. You might need to increase the arrival rate to see any issues.

Question 2: Does the model behave the way it was intended or are there animation issues still present?
________________________________________________________________________________

The patients will now wait properly in the waiting room. However Figure 14.4 shows an animation issue of a
patient waiting for a vet potentially in two places which did not occur in the previous models. In those
models, the SERVERS were the only constraining resource but suppose now we need a room (the SERVER) and a
vet (a RESOURCE). Since the input buffer is zero and we try to seize the vet (a resource) in the Before
Processing add-on process trigger. The entity can become stuck in a limbo state if the room (SERVER) is
available but the vet is not immediately available. Thus the entity cannot enter the input buffer because the vet
is unavailable, but yet it has seized capacity of the SERVER). The entity has not been completely transferred to
the processing station. An entity waiting for a station capacity (in this case the input buffer) to become
available will wait in an “Entry Queue”. Entry queues are used whenever an entity cannot enter a station
because there is no capacity available. The animation anomaly occurs because the entity is both waiting in the
entry queue and yet is in the station (due to the need for secondary resources).

Waiting to Enter
Server (Entry Queue) Partially Transferred
to the Processing
Station

Figure 14.4: Animation Anomalies

Step 15: We could change the Seize of the vet to occur in the “Processing” add-on process trigger which
happens after the entity has been completely transferred into the processing station. If you want to try it,
select SrvRoom3, remove the Before Processing209 add-on process trigger and then select the same the
process from the dropdown of the Processing add-on process trigger. Run the model and observe what
happens for the room.

208
You can select all four rooms and then set the Input Buffer to “0” under the Buffer Capacities properties at once. Here the
spreadsheet view is also useful.
209
You can remove an add-on process trigger either by using the right click reset menu option or selecting the trigger name and then
deleting it.
A Mobile Resource: The Worker 229

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 16: The animation anomaly has been removed but the patient now waits in the processing station for
the vet and then is processed when the vet arrives. However, this can be confusing as well as cause issues
with statistics since both waiting and processing are combined now. Table 14.2 describes the options for
choosing a MODELENTITY’s destination node from a list. The Selection Condition and Blocked Destination
Rule properties can be used to control which nodes are available for possible selection as the destination
node.

Table 14.2: Choosing Entity’s Destination via Select from List

Routing Logic Description
The goal determines the logic to utilize when selecting a destination node for an Entity.
This can be random or cyclic (i.e., will select the next node in the list), preferred (i.e., will
Selection Goal
always select the nodes that are first in the list unless a Selection Condition eliminates a
node), smallest and largest distance, and smallest and largest values.
Selection An expression evaluated for each candidate node destination when the Selection Goal
Expression property is set to the “Smallest or Largest Value.”
Selection An optional condition that is evaluated for each candidate destination node. This
Condition expression must evaluate to be true for the node to be considered under the Selection Goal.
Determines what happens if the destination nodes are blocked. The default value “Select
Available Only,” will only evaluate nodes to be selected that are currently not blocked
while “Select Any” will select any node regardless of whether it is blocked or not based on
Blocked the Selection Goal. The “Preferred Available”, will first select among all non-blocked
Destination Rule nodes and if all the nodes blocked will then revert back to select any. A node is considered
blocked if the node is an input node of an object and the number of entities assigned to the
station is greater than or equal to the capacity of the object (i.e.,
AssociatedStationLoad ≥ AssociatedStation.Capacity).

Table 14.3 looks at all possible scenarios associated with the server’s input buffer size and the values of the
Selection Condition and Blocked Destination Rule of the TRANSFERNODE. The “Modeling Result” column
represents what happens to the additional entities when the rooms all have one patient currently. The Selection
Condition will eliminate all nodes (i.e., filter the list) that do not meet the criteria before the selection goal is
used while the Blocked Destination Rule may not allow the node from being selected if the node is currently
blocked even though it was not filtered out.

Table 14.3: Understanding Blocking, Selection Condition, and Input Buffer Size has on Routing
Input Path TransferNode’s
TransferNode’s
Buffer Travel Blocked Modeling Result
Selection Condition
Cap. Capacity Destination Rule
1, All Entities will Wait at the SERVER’s
Infinity Any Rule None
Infinity Input Buffer.
Candidate.Node.
1, All Entities will wait at the
Infinity Any Rule AssociatedStationload
Infinity TRANSFERNODE (i.e., Waiting Room)
==0
1, Select Available All Entities will wait at the
0 None
Infinity Only TRANSFERNODE
Prefer Available, All Entities will wait at the SERVER’s
0 Infinity None
Select Any Input Node waiting to go into SERVER.
One additional Entity will wait at the
Prefer Available,
0 1 None SERVER’s Input Node waiting to go in all
Select Any
others will wait at the TRANSFERNODE.
Step 17: Set all the Input Buffer capacities back to “Infinity” or at least one to allow one entity to wait in the
room. Also making sure to reset SrvRoom3’s Before Processing and Processing add-on process triggers
back to SrvRoom3_BeforeProcessing and nothing respectively.
230 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 18: From Table 14.3, just relying on the blocking of the node’s associated station maybe problematic.
Therefore, we will set the Output@SrvWaitingRoom’s Selection Condition property to be
Candidate.Node.AssociatedStationLoad == 0 which says only allow nodes who currently
do not have any entities routing to the node or currently in the server to be eligible to be selected.

Figure 14.5: Forcing Patients to wait until a Room is Available

Question 3: For 40 hours, what is the average time in the system and the average wait in the room?
________________________________________________________________________________

Question 4: What assumption(s) does the current model make?

________________________________________________________________________________

Part 14.2: Using a WORKER as a Moveable Resource

The current model addresses the exam rooms, but does not model the veterinarians. A fixed RESOURCE object
similar to previous chapters could be used to constrain the processing to only one staff member, but the object
is stationary. Since the vet needs to travel between rooms, a dynamic moving resource would better reflect the
real system. WORKERS (and VEHICLES) in the system can be specified to be resources (i.e., they can be seized
and released like RESOURCES and have capacities like STATIONS) but they are visibly mobile.

Step 1: From the [Project Library], insert a new WORKER named WkrVet. Change the symbol of the
worker to an appropriate figure if you would like. Modify the properties as specified in Figure 14.6.

Change the Park While Busy property to True which will

force the worker to park when it is busy service a customer.
This will help minimize blocking issues on paths

Change the Initial Desired Speed to 4.5 kilometers per hour.

Specify that the Initial Node (Home) will be the office (i.e.,
TOffice) after performing Step 2 below.

Specify the Idle Action property to be “Park At Home”

Figure 14.6: Specifying the Worker Properties

Step 2: Next, build a separate network that the Veterinary staff will use to move throughout the system.
Add six TRANSFERNODES named TRoom1, TRoom2, TRoom3, TRoom4, TOffice, and TWaitRoom.
The TRoom nodes should be located near the same room with the office in the back as seen in Figure
14.7.

A Mobile Resource: The Worker 231

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 · Connect the nodes together using bi-directional PATHS since the staff can move in either direction to
service the next patient. Use logical distances of 15 meters for each link.210 The staff will travel on the
inner loop.

Vet
Network

Vet

Figure 14.7: Modified Vet Clinic with Vet Network

Step 3: Since the WORKER entity acts like a VEHICLE, the vets will utilize the parking queues when it stops
at each of the rooms and the office.211 By default, a parking queue is automatically placed. However, you
are unable to orient the queue or place it in a different location. If you wish to improve the animation, then
select one of the transfer nodes associated with the Vet network (TRooms or the TOffice node) and toggle
the Parking Queue button off. Next, select the ParkingStation.Contents Queue from the “Draw
Queue” section of “Attached Animation” ribbon and draw the animated queue very near the node selected
choosing the orientation. Repeat the process for the remaining nodes in the Vet network.

Step 4: Save and run the model and observe the dynamic worker.

Step 5: At this point our WkrVet stays in their office and does not actively seek any patients. The patients
will need to request the WkrVet in order to be processed. When the patient goes into processing they need
to seize the WkrVet and then release the WkrVet once they have finished processing.212

· From the Processes tab, create a new process named SiezeVet and insert a Seize step that will
request a WkrVet to service a patient as seen in Figure 14.8. Select all four room servers and
select the new process SeizeVet as the Before Processing Add-on process trigger.

210 Select all the paths and specify the direction and the logical length at one time. Note, careful consideration has to be followed in

using bi-directional paths because they can cause the system to become blocked and really should only be used when a very small
number of entities, workers or vehicles are being used. Otherwise drawing two paths (one forward and one backward) from each node
is the preferred method.
211
This is for animation purposes since the Vets would disappear when they arrive at their destination and begin service.
212
Again we will utilize the option that gives you the most control (i.e., processes). However, in this case Secondary Resources could
be used and is demonstrated in the commentary.
232 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 14.8: Seizing a Moveable Worker and Forcing a Visit
· Create a new Process called ReleaseVet. It will be triggered in the Add-on process trigger “After
Processing”. Insert a Release step that will release the Vet to service a new patient as seen in
Figure 14.9.

Figure 14.9: Releasing the Moveable Worker

· In the dropdown list for the “After Processing” Add-on process trigger for all four rooms, just
select the ReleaseVet to release the veterinarian after the processing has finished.
Step 6: Save and run the model and observe the veterinarian.

Question 5: Does the vet get seized and release correctly as well as move to the appropriate room?
________________________________________________________________________________

Step 7: Our vet is now being seized and released constraining the processing patients rather than assuming
four stationary vets at each room. However, the vet seems to be servicing the patients from the TOffice
node. We really do not want to seize the WkrVet worker until the vet is located at the room. From
Figure 14.8, you will notice the Request Move property which can be change to be “ToNode” and then
specify the appropriate destination node (e.g., TRoom1 for room one). This will force the WkrVet to
arrive at the TRoom1 transfer node before the processing can occur. The Move Priority property can be
used to specify the priority if multiple patients have requested the worker. Since this will be the same for
all four rooms except the node will change (i.e., TRoom1 to TRoom2), a tokenized process will be used.

· From the Definitions tab, add a new TOKEN named TknVet that will have a Node Object Reference
state variable named TStaWhichNode that will allow us to pass in the correct node as seen in Figure
14.10.

Figure 14.10: Creating a Specialized Token to Pass in the Move Node

· To create the Tokenized process, select the SeizeVet process from the Processes tab and specify the
TknVet as the Token Class Name property value as seen in Figure 14.11. Specify one input argument
that allow the user to pass into the process which room needs the service.

A Mobile Resource: The Worker 233

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 14.11: Making the SeizeVet Process Tokenized by Passing in one Input Argument
· Next, modify the Seize step to specify the Request Move to a particular node and the Destination
Node will be the TknVet.TStaWhichNode. This will force the vet to move to a particular node
before processing can begin on the patient taking into account the travel time to the room.

Figure 14.12: Modify the Seize to force Vet to Move to the Specified Node
· Finally, on each of the rooms specify the appropriate Which Room input argument property for the
Before Processing add-on process trigger as seen in Figure 14.13 for SrvRoom1.

Figure 14.13: Specify the Correct Node Associated with each Room
Step 8: Save and run the model observing how the veterinarian dynamically moves to each room.

Question 6: For 40 hours, what is the average time in the system and the average wait in the waiting
room?
________________________________________________________________________________

Step 9: Add a second veterinary staff member by changing the Initial Number in the System property under
the Population property category to be two.213

Question 7: For 40 hours what is the average time in the system and the average wait in the waiting room?
________________________________________________________________________________

Question 8: Is the second veterinary staff member justified and why?

________________________________________________________________________________

213
Another option is to insert a new worker named VetB. Then a resource list would have to be created to allow entities to seize one
from the pool of resources. This modeling approach may be necessary if the two staff members can service different types of patients,
different processing times, have a different ordering or work schedule, etc.
234 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Part 14.3: Returning to the Office between Patients

One additional modeling concern is the veterinarians need to return to their office after each visit to complete
paper work on the current patient and look over the chart of the next patient. All this needs to be done before
returning to one of the rooms to service the next patient. So we need to force them to go back to the office and
then delay for a particular time before going back to service the next patient. Since resources respond to
requests to be allocated (i.e., patients trying to seize them), one can model the situation where an entity
requests the WkrVet at the office so they will return back to the office and process that entity. In the office,
the paperwork uniformly takes between four and fourteen minutes.

Step 1: Save the current model as a new model and set the number of available veterinaries back to one.

Step 2: Insert a SERVER named SrvOffice near the TOffice transfer node and connect it to a SINK called
SnkOfficeWork via a connector as seen in Figure 14.14. Specify the processing time in the SrvOffice to
be a Uniform(4,14) minutes.

Figure 14.14: Modeling the Paperwork

Step 3: Insert a new MODELENTITY called EntPaperWork. It will be this entity that is processed in the
office. Since we want the Vet to do the paperwork before seeing any other patient, let’s give the Initial
Priority of the EntPaperWork a value of “2” (recall that the default value is “1”). Then to insure that this
paperwork is given priority, change the Dynamic Selection Rule property of the WkrVet to “Largest
Value First”, with the Value Expression as Candidate.Entity.Priority.214

Figure 14.15: Changing the Dynamic Selection Rule of the Worker

Step 4: When the vet finishes at a room, paperwork needs to be created and sent to the office. This
paperwork needs to Seize the WkrVet before the vet is seized by another room. Also as a part of this
process, the vet needs to be released from the room to visit the office. Therefore, we need to modify the
ReleaseVet process to accomplish all this, as shown in Figure 14.16. Recall that this process is used by all
the rooms when patients exit.

214
The Dynamic Selection Rule is used here only to illustrate its use and is strictly not necessary. The only case when the dynamic
selection rule is needed is when the Value may be changed while the entity is waiting, such as when a due date is reached.
A Mobile Resource: The Worker 235

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 14.16: Creating and Transferring the Paperwork
· Use the Create step to generate a paperwork entity a new TOKEN associated with the new entity.

Figure 14.17: Create Step to Produce an MODELENTITY

· Use the Transfer step to send the paperwork entity to the input buffer station of the office
SERVER.215

Figure 14.18: Using the Transfer Step to move the New Paper ENTITY.
Step 5: For the SrvOffice, create two new add-on process triggers processes for Before Processing and
After Processing.

Figure 14.19: Requesting and Freeing the Vet from Paper Work at the Office
· Copy the Seize step from one of the four rooms to request the WkrVet for the paperwork.
Remember to change the node to visit to the Toffice node in the Seize step.
· Copy the Release step from ReleaseVet to release the Vet after the paperwork is finished.
Step 6: Save and run the model and observe what happens.

Question 9: Is the model working as expected or do you observe any issues?

________________________________________________________________________________

215
Note this time we are sending the entity directly to the input buffer station rather than the Input node at the server which will force
the paper work to seize the vet.
236 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 7: The issue seems that the vet will process a second patient before heading back to the office.
Essentially, the vet processes the previous patient’s paper work but it is difficult to determine which paper
work went with which patient since all the patients and paper work entities are green. To alleviate the
situation, add three additional symbols for both the EntPatients and EntPaperWork giving each one of
the additional symbols a different color making sure the symbols for each entity type are the same color
(i.e., 0th symbol is green for both patient and paperwork, 1st symbol is magenta for both, 2nd symbol is
blue,and the 3rd symbol is red).

Step 8: Patients will be assigned a color based on the room they are serviced too (i.e., patients will be green
in Room 1, magenta in Room 2, blue in Room 3, and finally red in Room 4). Since this will be done when
the patients goes into processing, we need to add an additional input argument to the SeizeVet process to
specify which picture to use. First, add an INTEGER STATE variable named TstaRoomNumber to our
TknVet and then specify the second input argument of the SeizeVet process to be the room number as
seen in Figure 14.20.

Figure 14.20: Adding a new Input Argument to the Tokenized SeizeVet Process
Step 9: Next, add an Assign step to SrvRoom2_BeforeProcessing before the Seize step that assigns a
new value of “1” to the ModelEntity.Picture state variable.

Figure 14.21: Changing the Patient Color based on the Room

Step 10: Modify the Before_Processing for each of the rooms and specify “1”, “2”, “3”, and “4” for the
Room Number input argument for SrvRoom1, SrvRoom2, SrvRoom3, and SrvRoom4 respectively.

Figure 14.22: Passing the Room Number for SrvRoom1

Question 10: Why are we subtracting one from the room number in the Assign step?
________________________________________________________________________________
A Mobile Resource: The Worker 237

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 11: Save and run the model observing the entities as they enter the various rooms.

Step 12: Now, the patient entities are changing color based on the room they enter. The paperwork entity
now needs to reflect the patient color that created it. Process steps like Create and Search cause new
tokens to be created that are now associated with either the new entity created in the Create step or the
objects found in the Search step. Therefore, state variables associated with the original token’s object
(i.e., the ModelEntity.Picture) are not accessible to the new objects. One common approach to
copy states from one entity to another is to create a global (i.e., MODEL) state variable that is accessible by
all objects of the model.216 From the Definitions tab, add a new INTEGER DISCRETE STATE variable
named GStaPictureID.

Step 13: Next add an Assign step before the Create step in the ReleaseVet process that assigns the
GStaPictureID the ModelEntity.Picture of the patient entity that is creating the paper work. Then
insert another Assign step before the Transfer which is associated with the newly created paper work
entity that assigns the model state variable GStaPictureID to the ModelEntity.Picture state
variable as seen in Figure 14.23.

Figure 14.23: Copying the Value from One ModelEntity to another ModelEntity
Step 14: Save and run the model observing the entities as they enter the various rooms and the paper work.

Question 11: Can you see the different paperwork entities being created and how the vet always lags behind
by one?
________________________________________________________________________________

Part 14.4: Zero-Time Events

The issue seems to be that the vet essentially processes the previous patient’s seize even though the paper
work has a higher priority. A discrete-event simulation executes the simulation through an event calendar that
is ordered by the time of the event. Zero time events are events that occur at the exact same time. 217 The
Seize step for the paperwork entity happens at the same time the Release step occurs for the room. If the
release step occurs before the seize step occurs, then the vet may choose to service a patient in a different

216
Instead of creating a new Object in the Create step, a copy of the associated object could be created which would inherit all the
properties and states of the creating object. This approach could have been used in this case but the statistics associated with
EntPatients could not be used since this would include the paper work entities as well since they would be the same entity type. Also
we could have created a tokenized ReleaseVet process that would take in the room number and therefore would know the color that
needed to be set.
217
There are dozens of events that can occur that same time and can become an issue especially when dealing with allocations of
resources, workers, and vehicles.
238 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 room. Ordering of zero-time events can become critical in the execution of the model and the modeler won’t
know which is done first without some detective work.

A great tool for understanding what is happening during your simulation is through the use of the “Trace”.
The “Model Trace,” initiated from the “Run” tab, is the most complete description of what is happening.
However, in this case you want to trace the behavior of a process. You do this by selecting the step in the
Processes tab and clicking the breakpoint button in the ribbon, which will place the break point symbol
adjacent to the step218. Now in the “Facility” tab, select the Model Trace button which displays the trace in its
own window.

Step 1: Click on the Create step in the ReleaseVet process and select the breakpoint button. The
Create step should now have the breakpoint symbol beside it.

Step 2: Click the Model Trace button on the Run tab and then run the model via the Fast-Forward button.

Step 3: After the simulation “breaks”, click the Step button a couple of times. (You can recognize the break
because the line colored red describes the breakpoint.) When you click the Step button, a series of actions
are shown, colored in yellow, representing that simulation step. You may need to scroll back up to see the
breakpoint statement.

Step 4: Figure 14.24219 is a printout of all the events that have occurred around the breakpoint.

Figure 14.24: Sample Trace showing the Ordering of the Zero-Time Events
Step 5: Now you should read down the trace, noting the entity, the object, the step, and the action. Here
we see that the paperwork entity is created (EntPaperwork.63), the picture variable is assigned the value
of 0 and that entity is transferred to the “SrvOffice.InputBuffer.” Next, the patient entity (Entpatients.61)
releases one unit of capacity of the vet. Scrolling down through the other zero time events, you will see
that the vet will then respond to a request from another room first because the paper work entity does not
request one unit of capacity from the vet until near the very end of the current time. 220 All this is done at
the same simulation time but the ordering is important.

Question 12: Can you now see that the release step is done before the seize step of the office?
_______________________________________________________________________________

218
You can also right-click the step and select the “breakpoint”.
219
The trace is automatically saved to a CSV file as the model is running and can be imported into Excel for further analysis and
manipulation if necessary. You can also filter each column by selecting a row and the right clicking on the column heading.
220
It may be necessary to click the Step button again to get the next set of zero time events owing to the number of events that occur.
A Mobile Resource: The Worker 239

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 6: We need the release to happen after the Seize step which implies the concurrent time events need
to be reordered. To accomplish a reordering of events you can insert a Delay step before the release of
the vet (see Figure 14.25) that will delay the current token (representing the patient) a very small amount
of time, by specifying the delay time to be Math.Epsilon.221

Figure 14.25: Reordering Zero Time Events by Delaying an Epsilon Amount of Time
Step 7: Save and run the model again preforming a few “Steps” after the breakpoint which yields the trace
in Figure 14.26. As you can see the delay causes all the steps following it to become late priority events
meaning they will be placed at the end of the current time allowing the paper work to seize the vet first
which then allows the vet to select it based on its priority.

Figure 14.26: Reordered zero-time events

Step 8: Remove the break point by selecting the Create step and toggling the Breakpoint button. Now
run the model and observe what happens.

Question 13: Is the model working as expected?

_______________________________________________________________________________

Part 14.5: Handling Multiple Vets

In the previous chapter, there were multiple vets that could process patients. Like other dynamic objects
MODELENTITIES and VEHICLES, multiple WORKERS can be created at the start of the simulation or during the
simulation run. Let’s add two more vets to help in the clinic.

Step 1: Select the WkrVet and change the Initial Number in System property to “3” under the Population
category.

Step 2: Change the SrvOffice capacity to “Infinity” so each of the vets can service their own paperwork.

Step 3: Save and run the model observing the vets and paperwork.

221
Math.Epsilon represents the smallest real value number greater than zero. Delay steps will actually interpret the number as
zero, but it forces the event to be put at the end of the current event calendar. If the delay time is specified as zero, SIMIO will not
place the event on the calendar thus not changing the order at all. In the commentary, another solution to the ordering is presented.
240 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 14: Did the vet that processed the patient always process the correct paper work?
_______________________________________________________________________________

Step 4: The difficulty is the paper work entity just requested a vet and the closest one that was not busy
went to the office to perform paperwork (or was already sitting at the office) which was not what was
intended. We wanted the vet who performed the service to also complete the paper work. Therefore, the
paper work entity needs to request the same vet as the patient.222 First, select the MODELENTITY in the
[Navigation] panel and insert a new MODELENTITY DISCRETE INTEGER STATE variable named
EStaResourceID which will be used to store the ID of the resource to request for every paper work entity.

Step 5: Transition back to the Model and insert a new model DISCRETE INTEGER STATE variable named
GStaModelResID which will be used to pass the resource ID that was seized by the patient to the paper
work entity in a similar fashion as the picture number was passed.

Step 6: Next we need to set the global (i.e., model) variable (GStaModelResID) with the vet the current
patient has seized before we create the paper work entity and then set the paper work entity’s
EStaResourceID variable the value of model variable. Inside the ReleaseVet, modify the first Assign
step (i.e. GStaPictureID) to include another assignment which sets GStaModelResID a value of
ModelEntity.SeizedResources.LastItem.ID.223 Then, modify the second Assign step on
the “Created” path (i.e., PaperWorkColor) to set the ModelEntity.EStaResourceID variable a value of
GStaModelResID as seen in Figure 14.27.

Figure 14.27: Passing Seized Resource ID to the Newly Created PaperWork Entity
Step 7: Modify the Seize step in the SrvOffice_BeforeProcessing process to request a particular Vet
that needs to perform the paperwork based on the state variable ModelEntity.EStaResourceID. Under the
Advanced Options, specify the Selection Condition property to be Candidate.Worker.ID ==
ModelEntity.EStaResourceID as seen in Figure 14.28. Recall the CANDIDATE is a wild card
designator and will match all the WkrVets that are currently in the system. The Selection Condition
property has to evaluate to “True” for the WkrVet to be in the pool of potential ones to be seized and in
this example all WkrVets will be eliminated for consideration except the one whose ID matches the
EStaResourceID value.

222 All objects created in the simulation receive a unique identifier (ID) which can be used. For example EntPatients.61 is a patient

whose ID is 61.
223The vet will be the last resource seized since the patient had to first seize the server before it was allowed to go into processing and
then request the vet. See Table 13.2 for more information on the SeizedResources property of the entity.
A Mobile Resource: The Worker 241

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 14.28: Selecting Only the Correct Vet for the Paper Work
Step 8: Save and run the model observing the vets and paperwork.

Question 15: Did the vet that processed the patient always process the correct paper work?
________________________________________________________________________________

Part 14.6: Commentary

· Dynamic resources offer considerable modeling capability. In this chapter, we utilized processes to
seize and release the work stating it offers the most control over the process. However, in this case
we could have utilized secondary resources to accomplish the same task of the two processes used
to seize and release the vet. Figure 14.29 displays Resource for Processing property values under
the Secondary Resources section for the SrvRoom1 that would seize the vet and wait for it to
arrive before processing. However, notice the Selection Condition property used in the previous
section is not exposed. It is exposed for the Other Resource Seizes properties but using processes
was even easier.

Figure 14.29: Specifying the Seizing and Release of the Vet Using Secondary Resources
· We fixed the zero-time event by reordering the concurrent events using a delay step. Another
solution would have been to seize the vet before transferring as seen in Figure 14.30. The
“Original” token waits until the “Created” token reaches end of the current process. Therefore, the
seize will now happen before the release does in the zero-time events.

Figure 14.30: Seizing the Vet before Transferring the Paperwork

· Decision-making objects offer a number of modeling options especially as they relate to resources,
transporters, and workers. Because there is so much flexibility, it can confound the choices.
242 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 However, by working with these objects, you begin to develop some modeling patterns that seem
more appropriate than others.
· While it is our intent to use as many modeling approaches as we can, it will still be necessary for
you to supplement our efforts. Features such as the breakpoints, trace, and watch can be
extremely helpful in learning how SIMIO is behaving. It’s hard to modify a behavior unless you
understand the existing behavior.
· Unfortunately, the behavior of workers and vehicles is quite complex. Because of the complex
behavior, these objects could be referred to as “ heavyweight.” In some object-oriented designs,
heavyweight objects are composed of “lightweight” objects which are less complex. But that is not
the case with SIMSO, so we will continue to investigate and employ the complexity of these
objects.

A Mobile Resource: The Worker 243

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Chapter 15
Adding Detail to Service: A Bank Example
We have previously modeled a moveable resource as a worker. Entities would request the resource to visit a
specific node in the Seize step, which forces the worker to move to the node before allowing the entity to go
into service. VEHICLES can also be seized and used as resources in the exact same way. The WORKER object is
a specialized VEHICLE that can transport entities as well as being seized and released as a resource.224 The
modeling concepts relative to moveable resources used in this chapter will apply to vehicles or workers.

A small bank has al loan officer that assists customers with various loans and other account information as
seen in Figure 15.1. Customers arrive and wait for the available loan officer. The loan officer goes to the
waiting room and escorts (i.e., picks up) the customer back to the office to help the customer. After this
process is complete, some of the customers require a visit to their safety deposit boxes which then requires the
loan officer to escort them to the vault area and assist the customer again. The loan officer can be modeled
using a worker or a vehicle.

Banker and Customer

Customers Arrive 1.5 km p h
Exponentially(6.5) minutes Office Room
(10, 18, 20) minutes

55%
15 m
15 m
Bank Waiting Room
45%

5m

Safety Deposit
(10,15,25) minutes
15 m

Figure 15.1: Bank Example where Worker Picks up Customers as well as Services Them

Part 15.1: Using a Worker as a Resource and a Vehicle

One advantage of using a worker as a moveable resource is the worker can also be used as a vehicle. For
example, the vet or a technician may go to the waiting room and escort (i.e., picks up) the patient to the room
where the vet then sees the patient (i.e., vet is seized). That kind of transportation is being done in the bank.

Step 1: Create a new model.

Step 2: Insert a SOURCE named SrcWaitingRoom, a SERVER named SrvRoom, and a SINK named
SnkExit.
· Connect the output of the SrcWaitingRoom to the input of the SrvRoom via a 15 meter path and then
the output of the room to the SINK (SnkExit) by another 15 meter path.225
· Customers arrive exponentially with an interarrval time of 6.5 minutes.

224
A WORKER object must work “on demand” and has no routing type, meaning they don’t follow a specific sequence.
225
Select both paths and change the Drawn To Scale property to False and the Logical Length to be 15 meters simultaneously.
244 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 · The capacity of the SERVER should be one.
· Each customer takes between 10 and 20 minutes to be seen with most of them taking 18 minutes

Step 3: Insert a new MODELENTITY named EntCustomer and a WORKER226 named WkrBanker into the
model. The customers and banker travel at a constant rate of 1.5 km per hour or 1500 meters per hour.

· For the EntCustomer entity, add four additional symbols and change each of the five symbols to be
represented by one of the People\Animated symbols in the Library227. If you utilize just the “static”
“people” symbols, the people will “skate” across the paths. The animated symbols will allow the
people to physically walk, run, and/or skip. For the EntCustomer, set the Animation→Random
Symbol to “True” so it will randomly select one of the five people.
· Change the symbol of the WkrBanker to one that is more representative using one of the animated
people as well. To simplify the presentation, we delete all ten of the alternative symbol states.
· For the WORKER, move the animated ride queue (Ride Station Queue)228 so it is beside the WkrBanker
instead on top for a normal VEHICLE as seen in Figure 15.2. Notice that the ride station queue is
automatically given the option of “Match Attached Animation Speed” which means the “rider” will
walk at the same speed as the worker.
· Also for the worker, set the Initial Node (Home) to the Output@SrcWaitingRoom and set the
IdleAction to Park at Home.
· Now for the WkrBanker set the Park While Busy property under the Resource Logic section to
“True” to force the banker to park when it is utilized as resource.
· Make sure the SOURCE creates the EntCustomer MODELENTITIES.

Ride Station
Queue

Figure 15.2: Two and Three-Dimensional Versions of a Banker and Customer

Step 4: Save and run the model, making sure the entities are flowing correctly.

Question 1: Does it generate random people symbols arriving to the bank which are physically walking
and what is the banker currently doing?
________________________________________________________________________________

Step 5: In the next step, the WkrBanker should pick up the customers from the waiting room and deliver
them to the room. Change the outgoing transfer node (Output@SrcWaitingRoom) at the

226
One can also use a Vehicle as a transporter and resource.
227
Symbols in the Library can be selected based on “Domain”, “Type”, and “Action”.
228
By holding down the Shift key while in 3-D mode will allow you to move the queue up and down. Once you have it on the floor,
switch back to the 2-D view to position it beside the banker.
Adding Detail to Service: A Bank Example 245

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 SrcWaitingRoom to require a transporter as seen in Figure 15.3. Change the Ride On Transporter
property to “True” and choose a specific Transporter Type and specify the WkrBanker as the vehicle
requested.

Figure 15.3: Changing the Node Logic to Request a Transporter

Step 6: Make the path between the SrcWaitingRoom and the SrvRoom bidirectional, so the banker can
move between the two locations. Generally two unidirectional paths are better to prevent blockage.

Step 7: Save and run the model making sure the entities are moved by the WkrBanker to the room.

Question 2: What did you observe (i.e., does the WkrBanker stay with the EntCustomer while they are
being processed)?
________________________________________________________________________________

The WkrBanker is currently operating just as a VEHICLE which transports the customers from the waiting
room to the room. It does not necessarily stay with the customer while it is being processed, but continues
back to pick up the next customer and move them to the room.

Step 8: Insert a new TRANSFERNODE named TRoom229 above the SrvRoom and connect it via a one meter
bidirectional path to the input of the SrvRoom as seen in Figure 15.4.230

Figure 15.4: Model that Uses Vehicles as both a Resource and a Transporter
Step 9: In the Before Processing add-on process trigger of the SrvRoom SERVER, use a Seize step that
will seize the WkrBanker and request a visit to the TRoom node (see Figure 15.5 for more
information).231

229
This node will be used as the node to visit in the Seize block. Note, the node that a worker or vehicle is dropping off entities
cannot be the same node they need to visit. If the same node is specified, the model will result in a deadlock situation with the
Worker/Vehicle stuck at the node and the Entity frozen in the Inputbuffer and not able to proceed into processing.
230
This bidirectional path will not consume time because its speed is “Infinity”.
231Allocation as a Resource (i.e., responding to seizures) takes priority over handling ride requests from entities. If there is an entity
that has requested a ride and one that has requested its service as resource, the vehicle will handle the resource request first.
246 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 15.5: Requesting a Banker for a Customer to be seen at the TRoom Node.
Step 10: In the “After Processing” add-on process trigger of the SrvRoom SERVER, use a Release step
that releases the WkrBanker once it has seen the customer.232

Step 11: Save and run the model observing what happens.

Question 3: Did the WkrBanker transport the customer to the room and then move into place to service
the entity before heading back to get the next customer?
________________________________________________________________________________

Step 12: When the banker is seized by the customer at the office, it is automatically shown in the parking
queue of the TRoom TRANSFERNODE and the customer is shown automatically in the processing station
contents queue above the server SrvRoom. We will improve the animation to have the customer and
banker face each other across a desk as seen in Figure 15.6.

Parking Queue

Processing
Queue
Figure 15.6: Modifying the Animation of the Banker’s Office
· Select the SrvRoom server and the change the symbol to a desk by selecting the “Desk1” symbol in
the Library/Furniture section.
· Move the processing contents queue below the desk and change its alignment to Oriented Point. Make
sure the oriented points are facing the desk as seen in the figure.
· Next select the TRoom transfer node and turn off the automatic parking queue by toggling the
Parking Queue button in the Appearance→Attached Animation section. From the Draw Queue
dropdown, insert a Parking Station.Contents with Oriented Point alignment.233

232
We could have used the Secondary Resources section Resource for Processing that will seize and release the worker. However, in
the next sections we will make modifications that cannot be done with Secondary Resources.
233
You will need to orient the desk by selecting the desk. While holding down the Ctrl key, grab one of the corners and rotate the desk
around to the correct orientation.
Adding Detail to Service: A Bank Example 247

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 13: Save and run the model observing what happens when they arrive at the office.

Part 15.2: Having the Banker Escort the Customer to the Deposit Box

In this bank, 45 percent of customers need to visit their safety deposit box after seeing the banker.234 A similar
modeling method used previously to route patients to an X-ray machine with the assistance of the Vet will be
used. The WkrBanker resource will not be released if the customer needs to visit the safety box.

Step 1: Insert an additional SERVER named SrvSafetyDeposit.

· Set the capacity of the SERVER to one since there is only one vault area.
· The processing time for a customer to retrieve their safety deposit box and perform any actions
follows a Pert distribution with a min, most likely, and maximum values of 10, 15, and 25 minutes
respectively.
· Connect the output of the SrvRoom to the input of the SrvSafetyDeposit via a five meter PATH.
· Connect the output of the SrvSafetyDeposit to the SnkExit via a 15 meter logical PATH.

Step 2: Modify the SrvRoom_AfterProcessing add-on process trigger to model the banker behavior when
the customer needs to be escorted to the safety deposit box 45% of the time. From the Definitions tab, add
a new REAL PROPERTY named GProbSafetyDeposit and set the default to 0.45. Insert a Decide step (as
seen in Figure 15.7) that is probabilistic based with GProbSafetyDeposit as the expression.

Figure 15.7: Using Set Node to Send Customers to the Exit or Vault Machine
Step 3: If a visit to the safety deposit box is needed (i.e., the true branch), insert a Set Node step that will
set the destination node of the entity to the input at SrvSafetyDeposit.235 If not visiting the safety deposit
box (i.e., the false branch), release the WkrBanker and then set the destination node to the input at the
SnkExit.236 See Figure 15.8 for the property details of the two Set Node process steps. When the entity
enters the TRANSFERNODE of the SrvRoom where the Entity Destination Type has been set to continue, the
destination node that was set using the Set Node step will be used to route the entities.

234
The percentage is deliberately high to determine if the model is behaving correctly.
235
The Set Node step is identical to setting the Entity Destination property of a TransferNode (i.e., “Continue”, “Specific”,
“By Sequence”, or “Select from List”). The Destination Type can be a “Specific Node”, choose the node in the next sequence (i.e., “By
Sequence”) or a complicated expression.
236In the previous chapter, the SeizedResources functions were used in the link weights to route the entities to either the exit or
the X-ray machine.
248 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 15.8: Property Settings for the Set Node Property to Send the Entity to the Vault or the Exit
Step 4: Insert the “After_Processing” add-on process trigger for the SrvSafetyDeposit that will release the
WkrBanker once the service is completed as seen in Figure 15.9.237

Figure 15.9: Releasing the Banker after the Customer is finished at the Deposit Box
Step 5: Save and run the model.

Question 4: What is the advantage of defining a property instead of specifying 0.45 directly in the
Decide?
________________________________________________________________________________

Question 5: Do some of the customers flow to the safety deposit vault server?
________________________________________________________________________________

Question 6: Does the banker resource wait until the customer is completely done visiting the vault before
heading back to see another customer?
________________________________________________________________________________

Question 7: If the previous question is true, where does the Banker wait while the customer is visiting?
________________________________________________________________________________

Step 6: The WkrBanker does not move with (i.e., escort) the customer from the room to the safety deposit
machine. To facilitate the moving of the WkrBanker to the SrvSafetyDeposit, paths have to be added to
allow the WkrBanker to move to the vault room as seen Figure 15.10.

· Insert a new TRANSFERNODE named TSafetyDeposit near the SrvSafetyDeposit which will be the
node the WkrBanker will transfer to for processing the customer at the safety deposit machine.
· Insert a one meter path from the TRoom transfer node to the output node of the SrvRoom so the
WkrBanker can use the same path to escort the customer to the safety deposit box.
· Connect the input of the SrvSafetyDeposit to the TSafetyDeposit transfer node via a one meter PATH
and then connect the TSafetyDeposit node back to the output of the SrcWaitingRoom via a 15 meter
PATH which allows the WkrBanker to travel back to the waiting room to get the next customer.

237 Just copy the Release step from the SrvRoom_AfterProcessing to the SrvSafetyDeposit_ AfterProcessing.
Adding Detail to Service: A Bank Example 249

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 15.10: Model for Transferring Customers and the Banker
If the customer needs to see his safety deposit box, the WkrBanker needs to travel to accompany them to the
room. The process logic in the After_Processing add-on trigger in Figure 15.7 moves the entities to either the
safety box or the exit based on a probabilistic Decide branch using a Set Node process step. When this
process trigger runs, the TOKEN that is executing the process steps is associated with the entity. Therefore, any
steps (i.e., Transferring, Setting Nodes, etc.) are only applied to the entity238 associated with the TOKEN.
Step 7: Insert a Move step which will request a move of one or more moveable resources that have been
seized by the MODELENTITY (i.e., the associated object) as seen in Figure 15.11.

Figure 15.11: Using the Move Step to cause the Banker to Move to the Safety Deposit Box
Step 8: Save and run the model observing what happens with the banker and customer.
Question 8: Does the banker now move to the safety deposit box with the customer?
________________________________________________________________________________

Step 9: It may appear to move with the customer based on the speed of the animation. However, the banker
will actually move first to the TSafetyBox node before the customer does. To see this phenomenon, place
a break point on the path connecting the TRoom and Output@SrvRoom nodes. Fast forward the
simulation till it stops and then slow down the animation speed under the “Run” tab.

238 In some process steps, the action can be applied to the parent object of the Token’s associated object (e.g., the Server the logic
resides).
250 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 9: Does the banker move first now?
________________________________________________________________________________

Step 10: When the customer (i.e., MODELENTITY) requests a dynamic resource to be moved using the Move
step, the token associated with the entity will wait until the movement occurs before proceeding forward
(i.e., move itself). Therefore, we would like the banker and customer to move together. To manipulate the
banker, it will be necessary to identify the banker and then influence it. To accomplish that in this model
we will employ a Search process step. The Search process step can search a collection of items and
return Tokens associated with the items as specified in Table 15.1 and Table 15.2.
Table 15.1: The Properties of the Search Process Step
Property Description
Collection Type Type of the collection to search (See Table 15.2 for more information).
The type of search to perform (Forward, Backward, MinimizeReturnValue, and
Search Type
MaximizeReturnValue).239
Optional match condition that can be used to filter the items in the collection before
Match Condition
the search begins. Don’t forget to use the keyword CANDIDATE in the condition.
An expression that is evaluated for each item found by the search. The total sum for
Search Expression this expression is returned in the ReturnValue of the original TOKEN. One could
save the particular Resource and Seize the exact same one later.
One based index to start searching. By default it is one for Forward searches while the
Starting Index
number in the collection for Backward searches.
Ending Index One based index to end searching to assist in narrowing the search within collection.
An expression that specifies the maximum number of items that can be found and
Limit
returned.
Save Index Found Specify an optional discrete state variable to save the last item found.
Table 15.2: Different Types of Collections that Can Be Searched
Collection Type Description
Search all objects of a particular class (e.g., search all the entities that are currently
ObjectInstance
created)
EntityPopulation Search a particular entity population (i.e., MODELENTITY, WORKER, or TRANSPORTERS)
ObjectList Search the objects in a list (e.g., ResourceList, etc.)
NodeList Search all nodes in a particular list.
TransporterList Search all transporters in a particular list.
Search the objects currently seized by the Parent Object, Associated Object or a
SeizedResources
Specific Object
QueueState Search the objects currently in one of the queues
TableRows Search the rows in a Data table
NetworkLinks Search all links of a particular network.
NodeInboundLinks Search all the inbound links of a particular node.
NodeOutboundLinks Search all the out bound links of a particular node.
Step 11: If the customer needs to be escorted to the safety deposit box, insert a Search Step after the Set
Node as seen Figure 15.12. The “Original” branch is associated with the original TOKEN (i.e., customer)
while the “Found” branch will be invoked for a Token associated with each item found.

239
MinimizeReturnValue and MaximizeReturnValue search values will search the collection and return up to the limit of items
specified that minimizes or maximizes the Return Value expression property.
Adding Detail to Service: A Bank Example 251

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 15.12: Searching for and Transferring the Banker
· Set the Collection Type property to SeizedResources and make sure the Owner Type is
“AssociatedObject” which will search all the objects that are seized by the customer.
· Set the Match Condition to Candidate.Object.Is.WkrBanker to only search for the banker
that has been seized.240 This condition will insure that only the seized banker is.
· Set the Search Expression property to return the “ID” of the WORKER seized, which can be used for a
further enhancement.

Figure 15.13: Properties of the Search Process Step

Step 12: Once the WkrBanker has been found, we need to set the node to the TSafetyBox and then using a
Transfer241 process step transfer out of the parking station (i.e., the current stations) and into the
TRoom node as seen in Figure 15.12 with the properties as in Figure 15.14.242

Figure 15.14: Transfer Process Step Properties to Move Banker to the Vault
Step 13: Save and run the model.

Question 10: Does the banker now move with the customer when it needs to visit the safety deposit box?
________________________________________________________________________________

Question 11: Does it behave the way you envisioned?

________________________________________________________________________________

240
The customer entity has also seized one unit of capacity from the Server and we do not want to return the Server object.
241
The Transfer step can be used to transfer dynamic objects (Agents, Entities, Vehicles, etc.) from Current Node, Current
Station, Free Space, or Link to a specific Node, another station, Parents External Node, or outbound link.
242
If you transfer it directly to the TSafetyBox it will beam the banker there like Star Trek and not route it as this will take it out of the
station and place in the Troom node which then cause the banker to move to the TSafetyBox node.
252 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Part 15.3: Using the Transport Function of the Worker

The difficulty of the previous model is that the customer and the banker worker moved independently of one
another. Instead, the banker should pick up the customer (i.e., act as a vehicle) and then transport the customer
to the safety deposit vault where the customer would then Seize it again similar to moving the customer to
the banker’s office. Here the banker will both serve the customer and provide transportation.

Step 1: Save the current project.

Step 2: For the SrvSafetyDeposit SERVER, insert a Before_Processing add-on process trigger with a
Seize step (or you can use the Secondary Resources) to seize the WkrBanker once the customer is
dropped off. Use the Seize step in the Before_Processing that will seize the WkrBanker and request a
move to the TSafetyDeposit node as seen in Figure 15.15.

Figure 15.15: Seize the Banker at the Safety Deposit Vault

Step 3: However, only certain entities need to ride on a worker/vehicle (i.e., the ones heading to the safety
deposit box) while the other customers will walk to the exit. Therefore, the output node at the SrvRoom
cannot be specified to “Ride on Transporter” like the output node for the SOURCE since this would force all
customers to ride. A Ride process step can be used to force an “Entity to Ride” on a vehicle. However,
the Ride process step can only be utilized in a process trigger inside a TransferNode (e.g.,
Output@SrvRoom). Therefore, all of the logic used to determine where the entity will proceed after the
processing will need to be removed.

· Remove the After Processing add-on process trigger from the SrvRoom by choosing “null” from the
drop down243 or using “reset” on the right-click, since this logic will be needed in a different object’s
add-on triggers as seen in Figure 15.16.

Figure 15.16: Removing an Add-on Process Trigger from an Object

· Select the output TRANSFERNODE for the SrvRoom and then choose the
SrvRoom_AfterProcessing 244 process as the “Entered” add-on process trigger as seen in Figure
15.17. The Entered add-on process will be fired when an entity, worker, or vehicle enters the
TRANSFERNODE.

243You can also remove the process trigger by selecting it and then deleting it. Removing the process trigger only removes the process
from being invoked and does physically remove it
244 The SrvRoom_Processed is just the name of the process and can be easily changed. You could have created a new process trigger

and then copied all the logic into the new process.
Adding Detail to Service: A Bank Example 253

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 15.17: Specifying the Entered Add-On Process Trigger
Step 4: If the customer needs to visit the vault, the customer should request a ride and then release the
WkrBanker so it can then respond to ride requests.

· In the SrvRoom_AfterProcessing process, delete the Search step since it will not be needed.
· Insert a Ride step with the Transporter Type set to “Specific” and the specific Transporter Name
specified as a WkrBanker (see Figure 15.18).
· If the entity needs to see their safety deposit box and the customer has requested a ride, the
WkrBanker needs to be released so it can pick up the customer to transport it to the safety deposit
box. To reuse the same Release in the “False” branch, drag the end of the branch ( ) associated the
Ride step and drop it on top of the Release step.245

Figure 15.18: Adding the Ride Step and then Releasing the Banker
Step 5: Save and run the model.

Question 12: What do you observe when the customer needs to visit safety box now?
________________________________________________________________________________

Step 6: The banker is stalled at the desk while the customer waits for a ride to the safety deposit box. More
specifically the TOKEN associated with the entity in the process is waiting at the Ride step owing to the
Token Wait Action property (within the Advanced Options) which specifies, by default, to
“WaitUntilTranferredEvent” (i.e., until the entity has finished being transferred into the vehicle). Now
change the Token Wait Action property to “None (Continue)” as shown in Figure 15.19.

245 You could have copied and pasted another Release step after the Ride instead
254 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 15.19: Changing the Token Wait Action
Step 7: Save and run the model.

Question 13: What do you observe when the customer needs to visit the safety deposit box and the
WkrBanker moves to the output node of the SrvRoom?
________________________________________________________________________________

Question 14: Does it behave the way you envisioned?

________________________________________________________________________________

Step 8: When the WkrBanker, which is a VEHICLE, reaches the output node of SrvRoom to pick up the
customer, it actually enters the node thus running the same logic in Figure 15.18 which tells it to exit or
get picked up by a banker. This then creates the system deadlock. Only customers should run the logic.

Step 9: Insert another Decide step (see Figure 15.20) before the current Decide step and that checks to
see if the object associated with the Token is a MODELENTITY. The Is operator will check to see if the
object is a particular object which in this case Is.ModelEntity will return “True” if it is a customer
and “False” if it is Banker (i.e., WORKER/VEHICLE) as seen in Figure 15.20.

Figure 15.20: Using the Is Operator to determine if Object is an Entity

Step 10: Save and run the model.

Question 15: Does the model behave the way you envisioned?
________________________________________________________________________________

Step 11: The problem is the WkrBanker responds to ride requests in the order they arrive. The customers
going to the safety deposit machine should have priority.

Adding Detail to Service: A Bank Example 255

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 · Add an Assign step that changes the customer’s priority (ModelEntity.Priority) to two (see
Figure 15.21).
· Select WkrBanker and then change the Task Selection Strategy property under the Transport Logic
properties to be the “Largest Priority” as seen in Figure 15.22.246

Figure 15.21: Using the Entered Add-On Trigger to Specify a Ride on Vehicle

Figure 15.22: Changing the Task Selection Strategy of a Vehicle

Step 12: Save and run the model.

Question 16: Does the model work as intended now?

________________________________________________________________________________

Part 15.4: Resource Reservations

The previous section illustrated the use of a worker both for service and for transportation. However
continued use of this resource during a sequence of needs required a careful set of seizes, rides, and releases.
You should note that if another entity had requested a seize of the banker, that entity request would be handled
before responding to the ride request, regardless of the priority. In these cases, it is easier to use a resource
reservation. An entity may “reserve” a resource or worker or vehicle for continued use at different stages. By
reserving it, the entity may use it immediately or later. While being reserved, the resource cannot be used by
another entity. A resource may be reserved at a TRANSFERNODE or in the “Resource for Processing” in the
“Secondary Resources” section of an object or in a process using a Reserve step or the Release step.
Reservations may be implicitly cancelled using the Release step or explicitly using the Unreserve step.
Reservations made through the Secondary Resources will be automatically cancelled when service is
complete. So each customer arriving to the bank will reserve the banker for use throughout the process until
the banker is no longer needed.

Step 1: At the Output@SrcWaitingRoom, modify the “Transport Logic” to cause the transporter to be
reserved as shown in Figure 15.23. Adding this specification causes the banker to be reserved for this
customer and unavailable to any other customer for any type of service until the reservation is cancelled.247

246
Changing the Task Selection Strategy reorders the Rider Pickup Queue of the Vehicle.
247
The act of cancelling a reservation is called unreserving.
256 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 15.23: Creating a Reservation at Transport
Step 2: The customer requests service from the reserved banker at the office. After the banker serves at
SrvRoom, the customer will either release the banker and exit or keep the banker to be escorted to the
safety deposit. Change the SrvRoom_AfterProcessing to have a separate Release as shown in Figure
15.24 and the Assign step is no longer needed as we will continue reserving the WkrBanker.

Figure 15.24: Modifying the Process

Step 3: When the customer exits, the banker is released. Unless the Release step specifies keeping the
reservation, the reservation is automatically cancelled. However, those customers going to the safety
deposit box keeping the reservation in the Release step in the process going to the safety deposit box, as
shown in Figure 15.25. Note the Reservation Timeout expression property allows you specify logic that
will cancel the reservation if the entity still has reserved the object when the time expires.

Figure 15.25: Keeping Reservation on Release

Step 4: Rerun the model, looking closely at the behavior of the banker.

Question 17: Does the model work in a similar fashion as before?

________________________________________________________________________________

Part 15.5: Animated Entities

The “animated people” that have been added to SIMIO can enhance the visual appeal of a simulation. Each
animated people symbols (e.g., female, male, children, elderly, soldier, and cartoon people) has a list of
animation options. You can see the options by right-clicking an animated ModelEntity/Worker/Vehicle and
select “List Animations of Active Symbol”. Table 15.3 shows the partial lists for male and female animated
people. You can invoke an animation by its “Index” or its name “Name”. Unfortunately the animation lists are
not identical across people symbol groups as can be seen in the partial list as male animated people can

Adding Detail to Service: A Bank Example 257

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 perform 41 different animations while female animated people can perform 26 animations and not the same
ones as the males.

Table 15.3: Partial List of Animated Symbols for Male and Female Animated People
Male Female
Index Name Index Name
1 Walk 1 Walk
2 Run 2 Run
3 Jump 3 Jump
4 Attack_Kick 4 Scared

18 Walk_Stealth 18 Sit_Chair_Loop

41 Sit_Chair_Loop 26 Walk_Push
Step 1: To illustrate the use of animated entities, let’s reconsider the model from the first section of this
chapter. Add two chairs at the desk, one for the banker and one for the customer, using the Drawing →
Place Symbol, as shown in Figure 15.26. You may need to adjust the placements of the chairs, the parking
station queue and the processing station queue as you see the animation perform.

Figure 15.26: Add Chairs around Desk

Step 2: One of the MODELENTITY state variables that are defined by SIMIO in Definitions→ States is
“Animation.” This “string” state variable can be used to hold current entity animation and it can be used to
assign a new animated behavior. Specify the “Animation” properties section of the EntCustomer as
shown in Figure 15.27.

Figure 15.27: Animation Properties of the Customer

Step 3: WORKERS and VEHICLES do not have individual state variables like MODELENTITIES. So we need to
define a new Model DISCRETE STRING STATE variable named GStringStateWorker for the banker which
can be changed. So now specify the “Animation” properties of the WkrBanker as shown in Figure 15.28.

258 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 15.28: Animation Properties of the Banker
Step 4: The customer should walk to the banker’s desk and then sit in the chair. So in the Before
Processing add-on process trigger of the SrvRoom SERVER, add the assignment shown in Figure 15.29.

Figure 15.29: Customer Sits at Banker’s Desk

Step 5: Then the customer should walk to the exit or the safety deposit room. So in the In the After
Processing add-on process trigger of the SrvRoom SERVER, add the assignment shown in Figure 15.30.

Figure 15.30: Customer Walks to Exit

Step 6: The banker needs to sit at the desk. So in the in the “Entered” add-on process trigger of the
TRoom TRANSFERNODE, add the assignment shown in Figure 15.31.

Figure 15.31: Banker Sits

Step 7: The banker needs to walk again when they leave the desk to pick up the next customer or take the
current customer to the safety deposit box. So in the Exited add-on process trigger, add the Decide and
Assign step as shown in Figure 15.32. Recall we specified that the WrkBanker will park when they
are seized. When the WORKER object goes into the parking station, it exits the node and therefore would run
the Exited add-on process trigger forcing the banker to walk after they were told to sit. The IsParked
function returns true once the worker enters parking station before exiting the node and therefore will not
set the variable to “Walk.”248

248
The worker will enter the node when they leave the parking station as well executing the “Entered” add-on process trigger.
However, we do not need to add a Decide in the entered process because the Exited runs after the Entered making them walk.
Adding Detail to Service: A Bank Example 259

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 15.32: Banker Walks
Step 8: Save and run the model, looking carefully at the animation249.

Question 18: Are the animations working as expected?

________________________________________________________________________________

Part 15.6: Detailed Service: Tasks and Task Sequences

A service activity may be composed of several tasks rather than a simple processing time. For example, the
kitting of an electronic product may employ a sequence of tasks to complete the kit. Or an inspection may
require a number of tests at the inspection station. Often the detail in these types of operations can be
conveniently summarized by their processing time. There are, however, instances when the processing time
needs a composition of a set of interrelated tasks. SIMIO therefore recognizes two types of processing stations
within a SERVER, which are referred to as Process Types. Specifically, the alternative specifications are
“Specific Time” or “Task Sequence”. Currently, we have only modeled using the Specific Time which is the
default value. When more detail about a process needs to be included in the model, the model was extended
with more objects (i.e., additional servers were inserted). Using a Task Sequence process type, detail about
processing can be incorporated into a set of task sequences within a single SERVER.

Currently, the bank has a loan officer that assists customers with various loans, account information, special
deposits, and safety deposit box usage. A single processing time distribution was used to describe that service.
In the example, the loan officer escorted customers back to the office for general service. A portion of the
customers required a visit to their safety deposit box, which required the loan office to accompany them. In
that example we employed the WORKER for service and for transportation broke up the tasks into two separate
servers and flow. We want to re-consider the bank service and employ a different modeling approach to
incorporating details of the office service. More specifically, we want to recognize that the service being
provided by the loan officer is composed of a set of “tasks.” Each of the tasks has its own task properties.
Some tasks must be performed before others while some tasks may be performed in parallel at the same time.

Step 1: First, change the “Process Type” property of the SrvRoom to Task Sequence as seen in Figure
15.33.

249
Note that the TRoom and Input@SrvRoom nodes have two-way traffic and so the animation instructions are invoked in both
directions. In this case, that is not a problem, but it could be in other instances.
260 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Process Type

Figure 15.33: Changing Process Type

Step 2: This change reveals a Processing Tasks property which is specified through the REPEATING
PROPERTY EDITOR that allows you to list the tasks. Add a task as shown in Figure 15.34. It is important
to note that: (1) the Processing Time for the task is the same as before, (2) we added the Name of the task,
and (3) we specified that this task needs the WkrBanker and it should be moved to TRoom node.

Figure 15.34: Adding a Task

Step 3: By default a Sequence Number of 10 is provided. We accepted the Branch Type of Always and
specified the Process Type as Specific Time. We will modify these specifications in later sections.

Step 4: Since we are specifying the use of the WkrBanker in the task, we can remove the
SrvRoom_BeforeProcessing process and the Release steps in the SrvRoom_AfterProcessing process.
Note that in Figure 15.34 the resource is Reserved by default specification of the task. This specification
of a Reservation Timeout of Math.Epsilon insures this resource is held until the end of this time step
event so continued use of the resource is given priority.

Question 19: Why is the resource reservation important in this problem? (consider the need to accompany
the customer to the safety deposit box)
________________________________________________________________________________

Step 5: Save and execute the model.

Question 20: Is the model executing as it did previously?

________________________________________________________________________________

Question 21: Did the use of a task make it easier to create this
model?
________________________________________________________________________________

Adding Detail to Service: A Bank Example 261

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Part 15.7: Using Task Sequences

The previous example doesn’t illustrate the unique task capability especially since we handled it with
processes earlier or could have used secondary resources. A task sequence refers to a network of tasks ordered
by precedence. For example, you need a nurse to take vitals and information, then you need a doctor, and
potentially followed by more nurse visits. This would be harder to handle using one SERVER. Let’s consider
the banking service example which has more complications. Suppose the Loan Officer service is really
composed of three distinct tasks: (1) Task1: a review of the application, which takes Pert(3,5,8) minutes, (2)
Task2: a request for records, which only happens in 20% of the cases, and (3) Task3: a final assessment of the
loan, which takes Pert(5,7,10) minutes. The request for records doesn’t require time from the Loan Officer but
is an automatic search that takes a Uniform(2,6) minutes. In other words Task1 and Task2 are done in parallel
but only Task1 requires time from the loan officer while Task2 only requires time. Task3 cannot start until
both Task1 and Task2 are completed. A task sequence diagram is shown in Figure 15.35.

10: Task1:
Review Application:
Pert(3,5,8) minutes
20: Task3:
Task Sequence Number Final Assessment:
Pert(5,7,10) minutes
10: Task2: (20%)
10
Request Records:
Uniform(2, 6) minutes

Figure 15.35: Task Sequence Diagram

The general sequence number formation is X.Y.Z. and so forth where X,Y, Z, etc. are whole numbers. One
only needs to define the X value (e.g., 10 and 20) which represents the root number of the sequence and helps
to determine the precedence of the task. The Y’s, Z’s, and other numbers, are optional integer suffixes
separated by periods that are added to the root sequence to define subsequences (e.g., 10.1 and 20.2). Table
15.4 discusses the rules associated of how the task sequence numbers define the precedence of tasks.

Table 15.4: Rules of Precedence Associated with Task Sequence Numbers

Rule Description
If the task sequence number has only a root number (e.g., 20), then all tasks which have lower
Rule 1 task root numbers regardless of suffixes must precede that task (e.g., 10, 10.1, 10.2, 10.1.1 all
precede task with sequence number 20).
If a task A has a sequence number consisting of the required root number and one or more
suffixes (e.g., 20.1, 20.2, or 20.1.1), then tasks with sequence root numbers that are less, must
Rule 2
precede task A provided they either have no suffixes (i.e., 10 would precede 20.1 or 20.1.1) or
match all the suffixes (e.g., 10.1 precedes 20.1, 10.2 precedes 20.2, 10.1.1 precedes 20.1.1, etc.).
If a task A has a sequence number that has a root number followed by at least one suffix (e.g.
10.1, 10.2 or 10.1.1 and 10.1.2), then all tasks with the same root number without any suffixes
Rule 3
must precede task A (e.g., 10 would precede both 10.1 and 10.2). However, a task with sequence
10.1 does not precede 10.1.1 or 10.1.2 as these will all be considered starting at the same time.
Note, one or more tasks may have the same root sequence number with or without suffixes. For
example, task A could have sequence number 20 and task B could have sequence number 20
Rule 4 which means these two tasks will start in parallel once the preceding tasks have finished. Starting
at the same time is true for tasks with the same root sequence numbers followed by any number
of suffixes (e.g., 10.1, 10.1, 10.1.1, or 10.1.2 sequence numbers all start at the same time).

In our problem there are two task sequence numbers 10, so they must be done before the task sequence
number 20. Identical task sequence numbers are done in parallel as specified by Rule 4 in Table 15.4 (i.e.,

262 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 these tasks begin at the same time) and we will start task associated with sequence number 20 until both
sequence number 10 tasks are complete as stated by Rule 1.

Step 1: For the SrvRoom, will need to modify the one task that is defined and then add two additional
tasks.
· Modify the current task to represent the application review process as seen in Figure 15.36 which
requires the WrkBanker to be moved to TRoom and has a sequence number of 10.

Figure 15.36: Review Application Task Information

· Recall the request records task only occurs 20% of the time as seen in Figure 15.37 and requires no
resources in the current scope of the model. Change the Branch Type property to “Probabilistic” which
is explained in Table 1.5. Again we will use a property to represent the 20% of time a request is
needed. Right click on the Condition or Probability property and specify to create a new property
named GProbRecordCheck.

Table 15.5: Task Branch Type in the Task Sequence Definitioin

Branch Type Description
Always Branch will be executed all of the time.
Branch will conditionally be executed based on the expression independent of any other
Conditional tasks (i.e., if the expression evaluates to true then the branch is executed otherwise it
will be cancelled).
Branch will be executed probabilistically based on the percentage value stated and is
dependent on the other probabilistically branches at the same level. Probabilistic
branching is mutual exclusive meaning only one branch or none of the probabilistic
branches will be executed. For example, consider a sequence with two tasks (e.g., task
Probabilistic A and task B) each with sequence number 20. If Branch Type is set to “Probabilistic”
for both task A and task B with Condition Or Probability values 0.25 and 0.4
respectively. Both task A and task B cannot be executed at the same time. The system
will execute task A 25% of the time, task B 40% of the time, and neither task 35% of
the time.250 Note the sum of all the probabilities at the same level cannot exceed 100%.

250
If you wanted both task A and task B to be executed independent of each other, then set the Branch Type to “Conditional” with
the Condition or Property value set to Random.Uniform(0,1) < .25 and Random.Uniform(0,1) < 0.4 respectively.
Adding Detail to Service: A Bank Example 263

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 15.37: Request Records Task Information
· For the last task, the loan assessment task requires the WrkBanker again as seen in Figure 15.38
making sure to set the Sequence Number to 20.

Figure 15.38: Assess Loan Task Information

· Once you close up the tasks, we need to make sure the GProbRecordCheck property is set to 0.2.
Either right click on the MODEL in the [Navigation] panel and select Properties or select somewhere in
the model to access the MODEL properties, under the Controls→General section set the value to 0.2.

Question 22: Why do we not have to request a move for the banker in this task?
________________________________________________________________________________

Step 2: Run the model.

Question 23: Is the model executing as it did previously?

________________________________________________________________________________

Part 15.8: Some Observations Concerning Tasks

The addition of tasks to SIMIO is rather new and their potential is still being explored. However when a
service activity needs detailed modeling the task approach offers some interesting possibilities. For example,
let’s assume the banker may have two primary sets of tasks: loan application (75% of the time) and other
service visits (25% of the time). Also, for the Assess Loan task, 20% of the customers have issues and take
longer to process. For our problem, we would like to model the more complicated set of activities to be
performed by the banker at his office as seen in the task sequence diagram in Figure 15.39.

264 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 20.1
Banker Tasks
Review Application
10.1:
30.1
Loan Service
BankerTasks
Probability:
Assess Loan
ProbofLoanApp
20.1 40
Banker Tasks Probability:
Request Records ProbSafetyDeposit

10.2:
Other Services
Probability:
(1 – ProbofLoanApp)

Figure 15.39: Single Task Sequence Diagram

In this task sequence, the sequence numbers will utilize a suffix (i.e., 10.1), to denote a subsequence. One
subsequence performs the loan application tasks and a second subsequence consists of only one task that
performs other services by the bank with both subsequences potentially utilizing the visit to the safety deposit
box. Even though for this set of tasks some can be handled in a similar fashion as the last section (i.e., basic
processing time and seizing of resources), we will demonstrate the use of the other process types as described
in Table 15.6 to show the flexibility by using processes and sub-models. We will use processes to model the
“Normal Service,” “Assess Loan” application and the “Safety Deposit” tasks and utilize a sub-model for the
“Request Records” task. For the “Review Application” task we will use the same Specific Time process type
as well as for 10.1 which will be a zero-processing time task needed to complement the probability of
performing normal services as explained in Table 15.6.
Table 15.6: Process Type Option Explanation
Process Type Description
Specific Time Expression representing the time needed for the task to complete.
Specify a process that will be executed when the task starts. The task will finish once the
Process Name
process has finished executing.
Allows you to create a different model within the current facility window. When the task
starts, an entity of a type that is specified will be created and sent to the starting node of
Sub-Model
the sub-model. You can save a reference to the original entity to pass information to the
created entity. The task finishes once the created entity is destroyed by the sub-model.

Step 1: From the Definitions→Properties tab, insert a new “model” Expression property named
GProbofLoanApp with a default value of 0.75 for the probability that a customer needs a loan versus
other services. Repeat the process for GProbLoanAppIssues for the percentage of customers who have
issues with their loan application setting the default to be 0.20.

Step 2: As mentioned, we will utilize a sub-model for the request records to represent this as being done in
a different department once it submitted by the banker. In a different portion of the model, insert a new
MODELENTITY named EntRecordSearch, a SERVER named SrvRecordSearch, and a SINK named
SnkFinishRecSearch as seen in Figure 15.40.

Figure 15.40: Submodel for Request Records Department

Adding Detail to Service: A Bank Example 265

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 · Connect the SrvRecordSearch to the SnkFinishRecSearch via a connector as we are not concerned
with travel time.
· For the SERVER, set the Processing Time property to be the same Random.Uniform(2,6) minutes
to represent the time the department needs to process the request for record search.

Step 3: For the three tasks (i.e., “Assess Loan,” “Other Service,” and “Saftey Deposit Visit”), we will
create the processes that are associated with each task. Even though we could easily use the Resource
Requirements to request the WkrBanker as before and use the processes, we will Seize and Release
the bank as part of the process because we can control when they occur, in case of future enhancements.

· Insert a new process named TaskLoanApplication_AssessLoan as seen in Figure 15.41. The Seize
step should request the banker to the TRoom as seen in Figure 15.5. The Decide step should be use
“ProbabilityBased” with the Condition Or Probability property expression sent to 1-
GProbLoanAppIssues. Use the same Random.Pert(5,7,10) minutes for the loan
assessment with loans with no issues and Random.Pert(10,12,15) minutes for the assessment
of loans with issues. Then release the banker making sure to reserve the resource as in Figure 15.25
and the set the destination node of the customer to Input@SnkExit so this customer will be routed to
the exit.

Keep Reserved is
True

Figure 15.41: Process for the Assess Loan Task

· For the other services task, insert a process named Task_OtherServices. Copy all the steps from the
previous process changing the delay time to Random.Pert(4,6,8) minutes.

Figure 15.42: Process for Other Services Performed by Banker

· Finally, for the task associated with the safety deposit. Copy the TaskOtherServices process and
rename it to Task_SafetyDepositBox changing the delay associated with getting the key for the safety
deposit box takes Random.Pert(2,4,6) minutes.

Figure 15.43: Process for Safety Deposit Box

Step 4: Since we have defined the task submodel and the task processes, now we need to implement the
task sequence based on Figure 15.39 as seen in Figure 15.44. We will need to modify the three current
task sequences.
266 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 15.44: Implementation of Task Sequence
· From Figure 15.39, there are two potential sequence paths that customers can take (i.e., loan
application or other services). A Branch Type of Probabilistic will be used to select one of these two
paths. Therefore, we need to insert a new sequence task (10.1) named “LoanCustomer” that takes
zero time and has a Condition Or Probability value equal to GProbofLoanApp that allows the
selection of that sequence path.

Figure 15.45: Task Select Loan Customer Path

· Insert a new task sequence (10.2) named “OtherServices” that has 1-GProbOfLoanApp probability
to ensure that the customer will select either the 10.1 or 10.2 sequence paths. Change the Process Type
to “ProcessName” and set the Process Name property to “TaskOtherServices.”

Figure 15.46: Task Associated with Other Services

· Modify the “Review Application” task by changing just the sequence number to 20.1 so the task will
be executed only if the previous 10.1 is executed since the Auto Cancel Trigger property cancels
successive sequential tasks when the earlier task is not done.
· Modify the “Request Records” task to execute the sub-model instead of using simple processing time
as seen in Figure 15.47 as well as changing the sequence number to 20.1.

Adding Detail to Service: A Bank Example 267

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 15.47: Modification of the Request Records Task to Execute Submodel
· Modify the “Assess Loan” task by changing the sequence number to 30.1251 and specifying the
Process Type be “Process Name” and specify the TaskLoanApplication_AssessLoan process. Also,
make sure to remove the Resource Requirements of the WkrBanker.
· Finally, insert a new task sequence with a sequence number equal to 40 and use the
Task_SafetyDepositBox as the process. Also make sure the Branch Type is “Probabilistic” with the
Condition Or Probability set to GProbSafetyDeposit because only a percentage of customers perform
this task.

Figure 15.48: Task Sequence for Assess Loan and Safety Deposit Box.
Step 5: Modify the SrvRoom_AfterProcessing as shown in Figure 15.49. Remove the Set Nodes because
those are done in the task sequences. Modify the second Decide step is Figure 15.49 based on the
condition ModelEntity.DestinationNode == Input@SrvSafetyDeposit.

Figure 15.49: SrvRoom_AfterProcessing

Question 24: Why could we not put the Ride step in the task sequences?
________________________________________________________________________________

Step 6: Save and run the model. You might want to place breakpoints on the SrvRecordSearch to be sure
it is executed as expected.

Question 25: Does it run as you expected?

________________________________________________________________________________

251
A subsequence number is needed so this task is sequential to 20.1. A sequence number of 30 would have 10.2 as a prior task.
268 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 26: Has the use of tasks and task sequences enhanced the modeling of the banker’s service?
________________________________________________________________________________

Part 15.9: Commentary

· Using the vehicle or the worker for both transportation and service causes the model to become
very expressive of the real system.
· Resource reservations make continued use of a resource convenient.
· The animated people show promise to greatly enhance the visual appeal of the animation.

· Tasks and task sequence promote a detailed view of a service activity and can incorporate a wide
range of behaviors.
· Tasks have two add-on process triggers (i.e., Starting Task and Finished Task), which we did not
illustrate. These processes can be used to record when tasks start and when they end to help
calculate critical paths, earliest finish, latest start, etc.

Adding Detail to Service: A Bank Example 269

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Chapter 16
Modeling of Call Centers
In previous chapters, workschedules, non-homogenous rate tables, and routing have already been extensively
used. This chapter will reinforce those concepts while demonstrating how to handle two of the most common
behavior of people in queues: “balking” and “reneging”.

A call center receives support calls from customers who purchase kitchen appliances. The call center is open
seven days a week from 8:00 am to 7:00 pm (i..e,11 hours). It is, however, staffed from 8:00 am to 8:00 pm,
which leaves an hour to finish calls that are in-process at 7pm. There are two lines for technician service and
customers are routed to each (see Figure 16.1) depending on whether or not the first line can answer the
customer questions.

10%

Two Experienced Technicians

First Line (10, 15, 28) Minutes
Service
Call Ends
Calls Arrive 90%

NHPP Rate Table First Line Service

(3, 8, 15) Minutes

Figure 16.1: Model of a Simple Call Center

There is information about the arrival rate of calls during the day. Calls for support arrive at different rates
(calls per hour) throughout the day and th following table (taken from historical data) displays how the rates
change from hour to hour during the 12 hour day. Note that no calls are taken during the 11th hour.

Table 16.1: Hourly Rate of Support Calls to the Center

Time Interval Hourly Rate
8am - 9am 35
9am - 10am 55
10am - 11am 70
11am -12pm 85
12pm - 1pm 95
1pm - 2pm 110
2pm - 3pm 90
3pm - 4pm 88
4pm - 5pm 60
5pm - 6pm 50
6pm - 7pm 30
7pm - 8pm 0

Calls are serviced in the order they are received at the call center and, we are told that ten support technicians
answer the calls. Apparently the time it takes to answer a call generally takes 8 minutes to answer a support
question, with the minimum being 3 minutes and the maximum being 15 minutes. The first line service
270 Simulation Modeling with SIMIO

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 technicians can completely answer ninety percent of the calls. So after working with the customer on the
original question, ten percent of the calls must be routed to more experienced technicians, of which there are
two technicians. The second line technicians end up taking about 15 minutes on the call, with a minimum of
10 and a maximum of 28 minutes.

Since the call center uses telephonic equipment, the calls are transferred automatically to the service centers.
And, to the best of our knowledge, once a call comes in, the customer stays on the line until it is processed
(i.e., currently there are no call abandonments).When we are able to talk with people at the call center, we find
staffing is not constant throughout the 12 hour day, but has there are a number of different schedules for the
workers, including part-time and full-time. In fact, the schedule of work is described by the following table.

Table 16.2: Number of First Line Staff per Time Period

Time Period Number of First Line Technicians
8am - 9am 5
9am - 10am 5
10am - 11am 8
11am -12pm 10
12pm - 1pm 10
1pm - 2pm 16
2pm - 3pm 16
3pm - 4pm 12
4pm - 5pm 8
5pm - 6pm 8
6pm - 7pm 8
7pm - 8pm 2

Part 16.1: Building the Simple Model

Step 1: Create a new model that has one SOURCE named SrcCallsArrive, a SINK named SnkCallsExit, and
two SERVERS named SrvFirstLineTech and SrvAdvancedTech as well as add a MODELENTITY named
EntCalls.

Figure 16.2: SIMIO Model of Simple Call Center

Step 2: Connect all of the objects via CONNECTORS since these are electronically switched calls. To handle
the case that 10% of the calls have to be routed to the advanced technical support line, specify a selection
weight of 10 for the connector linking SrvFirstLineTech to SrvAdvanceTech and a selection weight of
90 for the connector between the SrvFirstLineTech and SnkCallsExit.

Modeling of Call Center 271

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 3: From the Data→Rate Tables section, create a new RATE TABLE name RateCallTable that has 12
intervals with an interval size of one hour.252

Figure 16.3: Arrival Rate Table for the Call Center

Step 4: Return to the model and change the Arrival Mode property of the SrcCallsArrive source to “Time
Varying Arrival Rate” and specify the RateCallTable as the Rate Table property.

Figure 16.4: Specifying the Arrival Mode of the Call Arrivals SOURCE
Step 5: Select the SrvAdvanceTech server and change the Initial Capacity property to two and set the
Processing Time property to Random.Pert(10, 15, 28) to simulate two advanced technicians.

Step 6: Next, the work schedule of the first line technicians needs to be created. Recall, SIMIO already
defines a default work schedule and day pattern for you. Select the “StandardDay” pattern from the “Day
Patterns” tab under the Data→Schedules section and modify it to match the data in Figure 16.5.253 Let us
use the WORKSCHEDULE on the left as it specifies every hour separately versus the one on the right which
groups some time periods into one for the same values. We will need each time period in a subsequent
section.

252
We could have modeled the whole 24 hours with the other intervals values being set at zero. However, we changed the run length of
our simulation to only be 12 hours.
253
We find it easier to set the Start Time and then the duration which will automatically set the End Time.
272 Simulation Modeling with SIMIO

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 16.5: Work Schedule for the First Line Technicians.
Step 7: Return to the Model and select the SrvFirstLineTech and specify the new work schedule and the
processing time as seen in Figure 16.6.

Figure 16.6: Setting the Properties of the First Line Technician Server
Step 8: Even though we do not allow any more calls to come after 7:00 pm, how many people on average
will be in the queues at 8:00 pm when there are no more technicians to handle these calls. An OUTPUT
STATISTIC can perform this calculation at the end of the simulation run. From the Definitions→Elements,
insert a new OUTPUT STATISTIC named OutStatNumberofCallsLeft with the following Expression
property: EntCalls.Population.NumberInSystem.

Step 9: In the Run Setup, change the Starting Time to be 8:00 AM and the Ending Type should be 12
Hours. Save and run the model.

Question 1: What is the average customer time in the system and the number of calls still in the system at
the end of the simulation?
________________________________________________________________________________

Question 2: What is the average utilization of the first line and the experienced technicians?
________________________________________________________________________________

Question 3: What is the average holding time in minutes for both of the two server’s input buffer?
________________________________________________________________________________

Question 4: What is the average number of calls waiting in the two server’s input buffer?
________________________________________________________________________________

Step 10: Next insert a new EXPERIMENT and run 20 replications of the model.

Modeling of Call Center 273

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 5: What is the average customer time in the system and the number of calls still in the system at
the end of the simulation?
________________________________________________________________________________

Question 6: What is the average utilization of the first line and the experienced technicians?
________________________________________________________________________________

Question 7: What is the average holding time in minutes for both of the two server’s input buffer?
________________________________________________________________________________

Question 8: What is the average number of calls waiting in the two server’s input buffer?
________________________________________________________________________________

Part 16.2: Balking

Now you observe that people who may encounter waiting for service sometimes abandon that service. In
queuing theory this is called “balking.” Effectively a person who observes the queue to be “too long” simply
will not join. In our case, the wait time is proportional to the number in the queue, so we will model the
waiting time by using the number (or contents) in the queue. If there are 20 or fewer people waiting only 10
% will balk while 25% will balk when there are 21 to 30 people waiting. And finally, if there are more than
30 people waiting, the chance of balking is 50%.

10%

Two Experienced Technicians

First Line (10, 15, 28) Minutes
Service
Call Ends
Calls Arrive 90%
Balk % First Line Service
NHPP Rate Table
(3, 8, 15) Minutes

Call Balks

Figure 16.7: Call Center with Balking

Step 1: Insert a new SINK named SnkBalk below SnkCallsExit and connect the SrcCallsArrive to the
new SINK via a CONNECTOR.

Step 2: To model the changing balking percentage based on the input queue of the first line service, a
LOOKUP TABLE offers a great modeling mechanism.254 From the Data tab, create a new LOOKUP TABLE
named FuncLookupBalkPercent with the appropriate values as seen in Figure 16.8.

254
There are many ways to model the balking procedure. One could add three TRANSFERNODES with the first transfer based on the
number in the queue by setting the link weights (i.e., <= 10 , > 10 && <=20, >20 && <= 30, and > 30). Then set the appropriate link
weights based on the percentages from the TRANSFERNODES to the SrvFirstLineTech or the SnkBalk.
274 Simulation Modeling with SIMIO

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 16.8: Balking Percentage Lookup Table255
Step 3: Since the expression is very long, we will setup a function to make it easier as well as limit errors
from occurring. From the Definitions tab, insert a new FUNCTION from the Edit section named
FuncBalkPercent. Set the expression to return a value from the Lookup Table based on the number
waiting at the SrvFirstLineTech input buffer (i.e., LookupBalkPercent[SrvFirstLineTech.
InputBuffer.Contents.NumberWaiting]) as seen in Figure 16.9.

Figure 16.9: Utilizing a Function

Step 4: For the connector linking the SOURCE to the SERVER (SrvFirstLineTech) set the set the Selection
Weight property to 1-FuncBalkPercent and the Selection Weight property of the link connecting the
source to the balking sink to FuncBalkPercent.

Step 5: Save and rerun the experiment.

Question 9: What is the average customer time in the system and the number of calls still in the system?
________________________________________________________________________________

Question 10: What is the average utilization of the first line and the experienced technicians?
________________________________________________________________________________

Question 11: What is the average holding time in minutes for both of the two server’s input buffer?
________________________________________________________________________________

Question 12: What is the average number of calls waiting in the two server’s input buffer?
________________________________________________________________________________

255
It is not strictly necessary to include the values for 0 and for 200, since the table assumes the last value if the lookup is outside the
range of the table – a value less than or equal to 10 will be 0 and values equal to or greater than 31 will be 0.5.
Modeling of Call Center 275

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 13: How many calls on average balked during the 12 hour period and do you feel the current
system is handling enough of the calls?
________________________________________________________________________________

Part 16.3: Modeling Reneging of Customer Calls

Now we allow customers to renege (i.e., leave) from the queue for standard calls when they have waited for a
period of time. In other words, we will allow abandonment of calls to occur. More specifically, every
customer who enters the call queue will wait for at least ten minutes. However if after waiting in the queue for
ten minutes, the customer examines its place in the queue (the telephonic system informs them of this). If they
are not within the first five places from the front of the queue, the customer reneges and leaves the system. If
they are within the first five places of the front, then the customer will extend their stay in the queue for
another four minutes in hopes of getting served. If after the extended stay, they are still in line, then they will
also renege.

Step 1: Save the current model and then insert a new SINK named SnkRenege below the SnkBalk but do
not connect it to anything. This SINK will be used to collect statistics on those customers that renege.

Step 2: The modeling methodology to be used is when a call comes into the service queue of the first line
technician that a delay occurs for ten minutes and then we check to see if it is still in the queue. If it is in
the queue and the position is greater than five you renege otherwise you wait another four minutes to see if
you have been served. Select the SrvFirstLineTech SERVER and insert the Entered Add-on Process
Trigger.

Question 14: What would happen if we placed the delay and checked for reneging directly in this process?
________________________________________________________________________________

Step 3: The actual call MODELENTITY would wait ten minutes then perform the check before actually
entering the server to be placed into the queue or be processed. Recall the discussion of how a token
associated with the actual MODELENTITY invokes the process and executes the steps. In this case we need
one TOKEN to continue on through the process (i.e., enter the queue as well as be processed) while a second
TOKEN goes through the delay and potential removal of the customer from the queue. Figure 16.10
illustrates how an Execute step can be used to spawn a new TOKEN associated (i.e., blue one) with the
entity to execute another process. When the first TOKEN associated with the “Call 21” entity runs the
Execute step, a new 2nd TOKEN is created that will execute the steps in the new RenegeProcess. The first
token can be allowed to continue on allowing the call to enter the queue and be possibly be served.

Associated
First Token
with Call 21
Continues on to
st
1 Token
k Enter Queue

Call 21 Enters SrvFirstLineTech

Invokes SrvFirstLineTech_Entered
Add-on Process Trigger

Associated 2nd
with Call 21 Token

Figure 16.10: Show how Two TOKENS are Associated with the Same Call MODELENTITY
276 Simulation Modeling with SIMIO

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 4: From the Processes tab, create a new process named RenegeProcess. Under the Advanced
Options, set the On Associated Object Destroyed to EndProcess. Refer to Figure 16.11. If the
MODELENTITY that is executing the process is destroyed, we want the RenegeProcess to end.256 This will
be important has it will be executing in parallel while the actual entity moves through the network.

Figure 16.11: Ending the Process if the Entity has been Destroyed
Step 5: Insert an Execute step into the SrvFirstLineTech_Entered process as seen in Figure 16.12
which executes the RenegeProcess process. Change the Token Wait Action property to “None
(Continue)” which allows the TOKEN executing this step to continue on. By default it is set to wait until the
RenegeProcess finishes, which would not allow the customer to continue to obtain service until the
renege process ends (specify the Execute step as shown in Figure 16.12).

Figure 16.12: Executing the Renege Process but Allowing First Token to Continue
Step 6: From the Definitions tab, add an INTEGER DISCRETE STATE variable named GStaPosition which
will be used to store the position of the caller in the queue.

Step 7: Return to the “Processes” tab and inset a Delay step into the RenegeProcess”as seen in Figure
16.13 that delays the TOKEN for ten minutes.257

Figure 16.13: Process to Handle Reneging

Step 8: After the ten minute delay, we need check to see if the associated call is still in the queue. If it is in
the queue determine its position in the queue. Insert a Search step after the Delay step which will be
used to search the queue as well as determine its position.258 Change the Collection Type property to

256
Otherwise, the SERVER will be executing the process without the MODELENTITY and it will crash when performing the Search step.
257 Typically it is a good idea to create properties for values rather than hardcoding values which makes it easier to change later.
258 See Chapter 13 for a discussion on using the Search step to look through collections.
Modeling of Call Center 277

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 “QueueState” and the Queue State Name to SrvFirstLineTech.AllocationQueue. Specify the
GStaPosition state variable as the Save Index Found property so it will be assigned the index of the item
found. Since we want to search for the associated call in the queue, specify the Match Condition property
to be ModelEntity.ID == Candidate.ModelEntity.ID which will search the queue and
match the ID of the item in the queue with the associated entity ID as seen in Figure 16.14.

Figure 16.14: Specifying the Search Step Parameters to Search for the Call in the Service Queue
Step 9: If the TOKEN associated with the EntCalls MODELENTITY finds itself in the queue, then the
GStaPosition state variable needs to be checked to see if it is less than or equal to five. Therefore, insert a
Decide step that performs a conditional test.

Figure 16.15: Checking the Index of the Entity in the Queue

Step 10: If the position of the entity in the queue is greater than five (i.e., the False branch), then we will
remove the entity from the queue and transfer it to the SINK SnkRenege. Insert a Remove step on the
“False” branch that will remove the current call associated with the “Found” TOKEN from the
SrvFirstLineTech.AllocationQueue.259 You will need to type in the value for the Queue State
Name property directly if it does not appear in the list.

Step 11: A TOKEN associated with the removed item will exit out the “Removed” path. Therefore, insert a
Transfer step on the “Removed” path to move the entity from the current station to the SnkRenege.

Figure 16.16: Transferring the Reneging Call to the Appropriate Queue

259 Recall the Remove step can be used to remove entities from storage and allocation queues only.
278 Simulation Modeling with SIMIO

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 12: If the position in the queue is less than or equal to five then delay the TOKEN for an additional four
minutes and search again. Insert a Delay step on the “True” path that delays for four minutes.

Step 13: Copy the first Search step and place it after the four minute delay to check to see if the entity is
still in the queue.

Step 14: If the call is found in the queue (i.e., it is still waiting to be serviced), it will automatically renege.
Grab the (End) associated with “Found” path and drag it on top of the Remove step to utilize the same
Remove and Transfer steps as seen in Figure 16.13.

Step 15: For those people who have been waiting for over ten minutes to talk with a technician, we would
like to know on average what their position was after ten minutes. From the Definitions→Elements section
insert a new TALLY STATISTIC named TallyStatPosition. In the RenegeProcess, insert a Tally step
right before the Decide step.

Figure 16.17: Tracking Average Line Position

Step 16: Save and rerun the Experiment.

Question 15: On average, how many calls renege during the 12 hour simulation run?
________________________________________________________________________________

Question 16: What is the average position for people who have waited at least ten minutes?
________________________________________________________________________________

Question 17: What is the average time in the system for calls that don’t balk or renege?
________________________________________________________________________________

Question 18: What is the average time in the system for calls that renege?
________________________________________________________________________________

Part 16.4: Optimizing the Number of First Line Technicians

We would like to determine if we have enough first line technicians at the right times that minimize our costs
and wait times by our customers. Optimization with OptQuest™ was introduced earlier and will be used again.
There are times when the WORKSCHEDULES may not be flexible enough and you may need to resort to another
technique. Nevertheless at the end of the section we return to specifying properties for the WORKSCHEDULE.

Step 1: Save the current model to a new name.

Step 2: From the Definitions→Properties section, insert 12 STANDARD EXPRESSION PROPERTIES with the
names and values as seen Table 16.3.260 These properties will be optimized to determine the best number
to have during each time period where TimePeriod1 would correspond to 8-9 AM.

260
Create one property with the correct category, a name of TimePeriod, and then copy it changing the default value to the correct
values.
Modeling of Call Center 279

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Table 16.3: Expression Property for the Time Periods
Expression Property Name Default Value Category
TimePeriod1 5 Call Center
TimePeriod2 5 Call Center
TimePeriod3 8 Call Center
TimePeriod4 10 Call Center
TimePeriod5 10 Call Center
TimePeriod6 16 Call Center
TimePeriod7 16 Call Center
TimePeriod8 12 Call Center
TimePeriod9 8 Call Center
TimePeriod10 8 Call Center
TimePeriod11 8 Call Center
TimePeriod12 2 Call Center
Step 3: Next, insert a new DATA TABLE named TableSchedule under the Data→Tables section which will
become our own manual schedule table as seen in Figure 16.18.

Step 4: Add a Standard Expression property column named Capacity.

Step 5: Insert 12 rows with the value of the expression for each row to be the corresponding property
created above as seen in Figure 16.18.261

Figure 16.18: Creating our own Work Schedule Table

Step 6: Since we are creating our own schedule, a TIMER will be needed to signal the time to change the
capacity of the SrvFirstLineTech server at given intervals. From the Definitions→Elements section,
insert a TIMER named TimerSchedule with a Time Interval of one hour and Time Offset equal to zero.

Step 7: Under the Definitions→States section, insert a new DISCRETE INTEGER State variable named
GStaTimePeriod which will be used to set the current time period.

261
The easiest way is to utilize Excel to create the 12 rows. First, add TimePeriod1 and TimePeriod2 into two different cells. Then
select both of these cells and drag down letting excel create the other ten rows. Then copy and paste the Excel table into the
TableSchedule.
280 Simulation Modeling with SIMIO

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 8: Next, navigate to the Processes tab and create a new process named ChangeCapacity that
responds to the TimerSchedule.Event as its Trigger Event property as seen below.

Figure 16.19: Changing the Capacity on the Timer Event

Step 9: Insert an Assign step that does the following two assignments. The first assignment will assign
GStaTimePeriod the value of zero through eleven by using the Math.Remainder function.262 The next
assignment will assign the current capacity of the SrvFirstLineTech SERVER a value from the table
TableSchedule.263 The simulation will start at Run.TimeNow equal to zero and the reason we add one.

Figure 16.20: Assignments to Perform our Manual WorkSchedule

Step 10: Return the Facility tab and change the Capacity Type back to “Fixed” since we are performing the
manual work schedule.

Step 11: In order to visually see if our new manual schedule is working, select the SrvFirstLineTech and
from the Animation tab, insert a new STATUS LABEL with an expression of CurrentCapacity.

Figure 16.21: Inserting a Status to Watch the Current Capacity

Step 12: Save and run the model comparing the results to previous model.

Question 19: Does the capacity of the server change at each hour break?
________________________________________________________________________________

Question 20: Compare the results of the following questions to the previous model.

What is the average customer time in the system and the number of calls still in the system at the end
of the simulation?
________________________________________________________________________________
What is the average utilization of the first line and the experienced technicians?

262 In this example, the remainder function takes care of the fact that Run.TimeNow will reach 12 and therefore the remainder will

be zero and start over. If you are running 24 hours for more than one day then using Math.Remainder(StaTimePeriod, 24)
is the way to get the correct time period for each day.
263
Row indexes start at one and therefore we need to add one to the StaTimePeriod to get the correct period.
Modeling of Call Center 281

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 ________________________________________________________________________________
What is the average holding time in minutes for both of the two servers’ input buffer?
________________________________________________________________________________
What is the average number of calls waiting in the two server’s input buffer?
________________________________________________________________________________
How many calls on average balked during the 12 hour period and do you feel the current system is
handling enough of the calls?
________________________________________________________________________________

Step 13: In the assignment, we manually converted the current simulation time into a row index (i.e.,
GStaTimePeriod) which if we ran the simulation for 14 hours would have started over at row one and
two for the 13th and 14th hour. DATA TABLES in SIMIO can be specified as being Time-Indexed which
allows one to access a column value based on the current simulation time. Select the table properties of
the TableSchedule and specify under the Advanced Options that it is Time-Indexed. The starting time will
tell SIMIO how to convert the current simulation time into which row while the Interval Size determines
the length of the time bucket to map to each row in the data table.264

Figure 16.22: Making a Data Table Time-Indexed

There are two table functions associated with time-indexed data tables that can be used to retrieve column
values or the current row index as described in Table 16.4.

Table 16.4: New Data Table Functions for Accessing Time-Indexed Columns
Function Description
The TimeIndexedValue will return the value
of the column or property that has been specified
TableName.ColumnName.TimeIndexedValue
for the current simulation time (i.e.,
TimedIndexedValue)
This function will return the current row index
TableName.TimeIndexedRow
based on the current simulation time.
Step 14: From the Processes tab, modify the Assign step by removing the first assignment which set the
StaTimePeriod and then set the current capacity to TableSchedule.Capacity.
TimeIndexedValue as seen in Figure 16.23.

Figure 16.23: Using the TimeIndexedValue Function to set the Current Capacity

264
SIMIO is internally doing what we did to calculate the row index.
282 Simulation Modeling with SIMIO

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 15: In the Facility window, insert two more status labels with expressions set to
TableSchedule.Capacity.TimeIndexedValue and TableSchedule.TimeIndexedRow
to see how these values change overtime.

Step 16: Save and run the model comparing the results to previous model.

Question 21: Does the capacity of the server change at each hour break in a similar fashion?
________________________________________________________________________________

Question 22: Compare the results of the following questions to the previous model.

What is the average customer time in the system and the number of calls still in the system at the end
of the simulation?
________________________________________________________________________________
What is the average utilization of the first line and the experienced technicians?
________________________________________________________________________________
What is the average holding time in minutes for both of the two servers’ input buffer?
________________________________________________________________________________
What is the average number of calls waiting in the two server’s input buffer?
________________________________________________________________________________
How many calls on average balked during the 12 hour and is this a good system now?
________________________________________________________________________________

Step 17: Now that we have created a manual work schedule which can be optimized to demonstrate the use
of time-indexed tables, we will utilize the built-in WORKSCHEDULE. From the Data tab, modify the
WORKSCHEDULE to utilize all the time period properties for each of the values as seen in Figure 16.24.265

Figure 16.24: Using Properties for the Values of a WORKSCHEDULE

Step 18: Save and run the model comparing the results to previous model.

Part 16.5: Using the Financial Costs as the Optimizing Objective

Since the TABLE to perform the task of developing a manual WORKSCHEDULE has been set up, we can use
OptQuest™ to optimize the 12 time period properties. Our goal is now to minimize the costs associated
staffing the call center while balancing the wait time of our customers along with the amount of balking and
reneging that occurs to reflect customer dissatisfaction. SIMIO introduced ABC costing for many of the
objects under the “Financials” property section. This cost information can be associated with dynamic (i.e.,
MODELENTITIES, TRANSPORTERS, and WORKERS) as well as fixed objects (i.e., SERVERS, SINKS, RESOURCES,
etc.). Therefore, you do not have to manually capture this information using state variables. The following
tables explain the financial cost properties of fixed objects (see Table 16.6), vehicle and worker objects (Table
16.5), and model entities (see Table 16.7).

265
The easiest way to set the values is to copy them again from Excel which can be used to create the column.
Modeling of Call Center 283

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Table 16.5: Financial Properties Associated with Vehicle and Worker Objects
Property Sub Property Description
Parent Cost Center See Table 16.6 for explanation.
Capital Cost Per
One time occurrence cost to add this VEHICLE or WORKER.
Worker/Vehicle
Cost associated with loading/unloading and transporting an
Cost Per Rider
entity regardless of ride time.
Transport Costs
This is cost rate applied per entity based on the amount of time
Transport Cost Rate
the entity waits in the buffer.
The cost charged per unit time for any scheduled capacity of
Idle Cost Rate
the server that is not utilized.
Resource Costs This is a onetime use for every time the object is used
Cost Per Use
regardless of how long the object is used.
The cost charged per unit time for any scheduled capacity of
Usage Cost Rate
the VEHICLE/WORKER that is utilized.

Table 16.6: Financial Properties Associated with Fixed Objects

Property Sub Property Description
Parent Cost Center The cost center element that costs which are accrued to this
server are summed up into. If the property is left blank, then the
costs accrued are rolled up into the parent object containing the
server (i.e., the MODEL).
One time occurrence cost to build this fixed object which occurs
Capital Cost
at the beginning of the simulation
Costs associated with the buffers. This is a onetime use for every
Buffer Costs
entity that enters the buffer regardless of how long they wait in
(Input, MemberInput, Cost Per Use
the buffer (i.e., all entities will enter the input and output buffer
MemberOutput,
regardless if the only stay in zero minutes.
ParentInput, ParentOutput
Holding Cost This is cost rate applied per entity based on the amount of time
and Output Buffer)
Rate the entity waits in the buffer.
The cost charged per unit time for any scheduled capacity of the
Idle Cost Rate
Resource Costs object that is not utilized.
(All fixed objects) This is a onetime use for every time the object is used regardless
Cost Per Use
of how long the object is used.
Resource Costs (Except The cost charged per unit time for any scheduled capacity of the
Usage Cost Rate
WORKSTATIONS) object that is utilized.
The cost charged per unit time for any setup activity at a
Setup Cost Rateτ
WORKSTATION.
Resource Costs Processing Cost The cost charged per unit time for any processing activity at a
(WORKSTATIONS Only) Rateτ WORKSTATION.
Teardown Cost The cost charged per unit time for any tear down activity at a
Rateτ WORKSTATION.
Table 16.7: Financial Properties Associated with ModelEntities Objects
Property Description
Initial Cost Initial cost for the creation of the MODELENTITY.
The cost charged per unit time for any scheduled
Initial Cost Rate
capacity of the server that is not utilized.

Step 1: Select the SrvFirstLineTech SERVER and expand the “Financials” property category. The first line
technicians receive $25 per hour regardless if they are helping a customer or not. Therefore set the Idle
284 Simulation Modeling with SIMIO

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 and Usage Cost Rate properties to “25” as seen in Figure 16.25. Owing to the trunk line rentals, there is a
fixed cost (i.e., $10) regardless of usage as well as $5 per hour per call usage rate that needs to be
specified. We will leave the Parent Cost Center blank to utilize the MODEL’s cost center.266

Figure 16.25: Specifying Cost Information on the SERVER

Question 23: Save and run the model observing the new cost values in the results section. What is the total
cost of the system associated with the MODEL?
________________________________________________________________________________

Step 2: Since our cost information is setup, we now need to optimize the number of technicians by
minimizing the total costs associated with the system while balancing the number of unhappy customers.
Create a “New Experiment” named Optimization.

Step 3: Add the following three responses to the experiment that will determine the total cost of the system
as well as keep track of the number of people who balk and renege.

Figure 16.26: Add the Three Responses of Cost, Number Reneging, and Number Balking
Step 4: Like many problems, this is a multi-objective problem where the objectives often conflict (i.e., if
you run the optimization it will most likely chose only one technician at each time period while customers

266
A COST CENTER element can be specified to have the cost summed up. These cost centers can be specified in a hierarchical fashion
(i.e., several object roll up to one center for total costing (e.g., department costing), several centers can roll up to higher level centers
(e.g., departments to a plant, plants to organizations, etc.). You can also use Assign steps to assign costs to centers dynamically.
Modeling of Call Center 285

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 would like the maximum allowed at each period. As was done previously, we can convert the multi-
objective by combining the objectives in a weighted fashion, which generally is not usable idea. 267
Instead, we will utilize thresholds or bounds on the number of balking and reneging as seen in . Having 25
people balk or renege represents on average approximately 3% of the callers that are seen in day which is
what management feels is adequate for customer satisfaction. We set the upper bound of 15 and 10 for the
number of people it is acceptable to balk and renege respectively as seen in Figure 16.26.

Step 5: From the Add-Ins section of the new experiment, select the OptQuest for SIMIO add-in. Next, for
each of the 12 Controls variables, specify the Minimum Value to be “1”, Maximum Value to be “20” and
the Increment property value to be “1”. Properties of the model become Control variables automatically.

Step 6: Specify the TotalCost response as the Primary Response as seen in Figure 16.27. Also, set the
OptQuest™ parameter values based on the figure. We will need to run this optimization longer based on
the number of variables and threshold constraints.

Figure 16.27: OptQuest™ Parameter Values

Step 7: Save and run the optimization experiment.

Step 8: Once the optimization is complete filter the scenarios such that only the ones that are checked are
present which represents the scenarios that meet the number of balker and reneging thresholds.268 Next
sort the total cost column in ascending order. At this point you should run the K-N selection algorithm on
the top 20 or so scenarios to determine the best since only seven replications were used. Use a $100
indifference zone.

Question 24: What did you get for the total cost of the best system?
________________________________________________________________________________

Question 25: What did you get for the average number of balkers and renegers?
________________________________________________________________________________

Question 26: What were the number of technicians needed for each time period?
________________________________________________________________________________

Step 9: The optimization has determined that greater than 140 technician hours are needed based on costs.
However, management only has access to 135 total hours. Add a constraint which is a constraint on the
decision control variables that utilizes the following expression as seen in Figure 16.28.

TimePeriod1+TimePeriod2+TimePeriod3+TimePeriod4+TimePeriod5+TimePeriod6+
TimePeriod7+TimePeriod8+TimePeriod9+TimePeriod10+TimePeriod11+TimePeriod12

267
We could arbitrarily put a large cost on the input buffers of both the balk and renege SINKS to try to force the system to minimize
the number of balking and reneging that occurs. However, determining the right value of the cost can be difficult.
268
Select the filter option in the column heading row above the check boxes.
286 Simulation Modeling with SIMIO

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 16.28: Specifying Control Variable Constraint
Step 10: Save and run the optimization experiment.

Step 11: Once the optimization is complete filter the scenarios such that only the ones that are checked are
present which represents the scenarios that meet the number of balker and reneging thresholds. Next sort
the total cost column in ascending order.

Question 27: What did you get for the total cost of the best system as well as the average number of balkers
and renegers?
________________________________________________________________________________

Question 28: What were the number of technicians needed for each time period as well as the total number
of technicians needed?
________________________________________________________________________________

Step 12: Next, run the KN Select Best Scenario algorithm with an indifference zone of $10 on the feasible
solutions.

Question 29: What do you observe about the feasibility of the solutions now?

Step 13: Recall the KN selection algorithm does not take into account constraints on the non-primary
objectives. Select the minimum one that meets the two objective thresholds.

Question 30: What did you get for the total cost of the best system as well as the average number of balkers
and renegers?
________________________________________________________________________________

Question 31: What were the number of technicians needed for each time period as well as the total number
of technicians needed?
________________________________________________________________________________

Part 16.6: Commentary

· In this chapter, the Execute step was used to create a new TOKEN that worked independently of
the original TOKEN. There is an additional modeling concept that can cause the same phenomena to
occur. The Exited add-on process triggers for objects can work in a similar fashion. The Exited add-
on process triggers are executed when the trailing edge of the MODELENTITY has exited the node,
link, station, or object depending on the trigger. Many people do not realize that logic placed in
these add-on processes actually run in parallel while the entity continues on to the link, station, etc.
To illustrate this fact, you can clear the Entered Add-on process trigger for the SrvFirstLineTech
and then select the Input node of the server. Set the Exited add on-process trigger of the
Input@SrvFirstLineTech BASICNODE to be the RenegeProcess and rerun the model comparing
the results to the previous section.
· More complicated reneging processes can also be utilized. For example, every five minutes you are
told how many people are still in line and the person could make a probabilistic decision based on
some behavior (i.e., number of people and how long they have already been waiting) as seen in
Modeling of Call Center 287

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 16.29.In this example, if they search finds the entity in the queue as before, it will make a
decision to renege or not. If they decide not to renege (i.e., “True” path), it loops back and delays
for another five minutes. If the caller decides to renege, the same remove and transfer is utilized.
The looping will continue until either the caller no longer remains in the queue or the caller reneges
and leaves the system.

Figure 16.29: Reneging Process that Checks Every Five minutes

288 Simulation Modeling with SIMIO

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Chapter 17
Sub-Modeling: Cellular Manufacturing
Object-oriented simulation languages offer the ability for users to extend the base language by creating new
objects through composition and extension/inheritance. Most object oriented languages offer the ability to
create new objects out of existing ones which is referred to as composition. SIMIO provides the ability to
create objects via composition as well as through extension. This chapter will explore the easier object
creation (i.e., composition having facility models inside of other models (sub-models).

A manufacturing operation has ten multiple work cells where each work cell consists of two machines that are
connected by a short conveyor as seen in Figure 17.1. A dedicated operator is needed to load parts into the
machine for processing on the first machine and then to unload them from the second machine. The conveyor
system will automatically transfer the parts from the first machine to the next machine. Some of the work
centers (i.e., machines within the work cell) are capable of handling multiple parts in parallel. Also, priority is
given to unloading parts from the second machine to ensure parts are processed through the cell quickly. One
way to do this would be to insert ten Machine A servers, ten Machine B servers, ten conveyors, and ten fixed
resources that are seized and released. Since the work cells are the same we can use composition to create a
work cell object and then use it nine times to model the whole manufacturing system.

Work Cell 1
Work Cell 2
`
`
`
Work Cell 10

Machine A Machine B
Conveyor
A B

Operator Loads A and Unloads B

Priority for Unloading

Figure 17.1: WorkCell Manufacturing Facility

Part 17.1: Model of One Work Cell

Before building the larger model, it is better to work initially with a single work cell. Eventually, this model
will become a building block (i.e., composition object) and is referred to as a “sub-model” (see Figure 17.2).

Conveyor Speed of 2.5 meters per minute

Conveyor length = 5 meters
Parts arrive every 6 minutes

Parts Arrive Machine A Machine B Exit

10 Minutes 10 Minutes

Processing 4 minutes Processing 3 minutes

An Operator takes 1 minute to load Machine A and 30

seconds to unload Machine B

Figure 17.2: Individual Work Cell

Sub-Modeling: Cellular Manufacturing 289

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 1: Create a new project SIMIO. Change the name of the model to WorkCell by changing the Model
Name in the Model Properties.

Step 2: Construct the model according to Figure 17.3.

Figure 17.3: Initial Two-Machine Model

· Add two SERVERS named SrvMachineA and SrvMachineB. The processing times should be four
and three minutes respectively for now.
· Connect these two machines by a five meter conveyor with a speed of 2.5 meters per minute.
· Add a MODELENTITY, SINK and SOURCE that generates arrivals every six minutes. Do not worry
about naming since these are only temporary.
· Connect the SOURCE to SrvMachineA’s input and the sink to SrvMachineB’s output node via ten
minute time paths.
Step 3: Save and run the model to make sure entities flow correctly.

Step 4: The machines require an operator to load the part into the first machine which generally takes one
minute and then to unload the part from the second machine which takes approximately 30 seconds. Insert
a new fixed RESOURCE named ResOperator. Add an additional symbol to indicate working and color the
iris green.

Step 5: Add the “Before_Processing” add-on trigger for Machine A that will seize the “Specific” resource
ResOperator, delay for a certain amount time, and then release the specific resource. This process
represents parts that enter processing at Machine A but need to be first loaded into the machine by the
operator before being processed.

Figure 17.4: Process Logic to Seize and Release Operator to Load First Machine
Step 6: The Delay Time property for the Delay step should be a referenced property. Right click on the
Delay Time and “Create a New Reference Property” named LoadDelayTime.

Figure 17.5: Specifying a Referenced Property as the Delay Time

Step 7: Repeat the procedure for the unloading of parts out the second server SrvMachineB. The only
difference is the name of the referenced property should be UnloadDelayTime and the add-on trigger
should be the “Processed” trigger (i.e., SrvMachineB_Processed). Once the process has been created,

290 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 just copy and paste the process steps from the SrvMachineA_Processing and create the new reference
property for the Delay Time.

Step 8: Switch to the “Definitions” tab, change the Category of the two new properties to “Process Logic”
and set the Default Units property to “Minutes.” The default value should be set to one minute for the
LoadDelayTime property and 0.5 minutes for the UnloadDelayTime property.

Step 9: Save and run the model making sure it behaves as intended.

Question 1: For 24 hour replication, what is the average time in system (in minutes)?
________________________________________________________________________________

Part 17.2: Creating the Sub-Model

Currently, we have the basic work cell model working correctly but we need to create the sub-facility model.
In order to create the sub-model, an External view of the sub-model needs to be defined before you can use it
as an object in other models.

Step 1: Delete the SOURCE, SINK, and MODEL ENTITY from this facility model since they were only needed
to test the logic of the two machines connected via a conveyor.

Step 2: Since the WorkCell model will be used as a sub-model, none of the nodes in the model (i.e.,
Input@SrvMachineA or Output@SrvMachineB) are visible to external model (i.e., models using the
WORKCELL object). Therefore, two external nodes need to be added to the external view of the object (i.e.,
Definitions→External section) that facilitates the flow of entities into and out of the sub-model.269

· Since entities entering the WORKCELL object need to enter the input node of machine A (i.e.,
Input@SrvMachineA), right click the BASICNODE to bring up the sub menu seen in Figure 17.6 and
select “Bind to New External Input Node” and then name the external node Input to be consistent.270
· Next, right click the output node of SrvMachineB but this time select “Bind to New External Output
Node” and name it Output.
· By default, all objects (i.e., animated queues, objects, etc.) in the facility are visible to the external
world as indicated by the “Externally Visible” button being toggled on as seen in Figure 17.6. Toggle
the externally visible button off for both the Input@SrvMachineA and Output@SrvMachineB
nodes. This will make it easier to see our new Input and Output nodes.

Figure 17.6: Automatically Adding External Nodes as Making Objects Visible to the External View

269
All the basic objects (e.g., SOURCE, SERVER, WORKSTATION, etc.) have nodes in their external view that allow entities to flow into
and/or out of the object.
270
These can be added manually from the External panel underneath the “Definitions” table however all of the property settings also
have to be done manually.
Sub-Modeling: Cellular Manufacturing 291

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 3: Navigate to the External panel underneath the Definitions tab and survey the external view which
includes the two external nodes named Input (a BASICNODE) and Output (a TRANSFERNODE) as well as a
symbol representing the model as seen in Figure 17.7. Only the two nodes are accessible.

External External
Input Output

Figure 17.7: External View of the WorkCell Object

Step 4: Select the Input node which can be seen as an instance of the BASICNODE NODE Class. Generally
entities are transferred from this input node into a station. However, the sub-facility model does not
expose a station. Therefore, these entities will be transferred to the input of the SrvMachineA. The Input
Location Type property is set to “FacilityNode” and the node is the Input@SrvMachineA as in Figure
17.8. As entities flow into the Input of the WorkCell they will be automatically routed to SrvMachineA.

Figure 17.8: The Input Node

Step 5: After selecting the Output node, you can see this is an instance of the TRANSFERNODE class to
allow entities to be routed to their next destination once they leave the WorkCell (see Figure 17.9).

Figure 17.9: The Output Node

Step 6: Entities that enter the external input node of the WORKCELL object will automatically be transferred
to the input of the SrvMachineA via the Input node. When entities reach the output node of the
SrvMachineB, they need to be transferred to the external output node (Output) defined in the external
view. Switch back to the “Facility” tab and select the Output node of SrvMachineB (i.e.,
Output@SrvMachineB). Under the Advanced Options category, you can see that SIMIO automatically
changed the Bound External Output property to specify the Output as the external node name property as
shown in Figure 17.10. Now entities will automatically be transferred to the external Output when they
reach the output of the SrvMachineB.

292 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 17.10: Connecting the External Node
Step 7: Save the WorkCell model.

Part 17.3: Creating a Model using the WorkCell Sub-model

The WORKCELL Sub-facility model has been built and can now be saved to a library and used as any other
model that is part of SIMIO. We are going to build our manufacturing systems using the WORKCELL.

Step 1: From the Project Home tab, insert/create a new “Fixed Class” model.

Step 2: In the new model, insert a SOURCE named SrcParts, a SINK named SnkExit, and new work cell
named Cell1 which is part of the Project Library panel. Connect the SOURCE and the SINK to the
WORKCELL via a ten minute time paths as seen in Figure 17.11.

Step 3: Make the interarrival time of parts be Exponential with a mean of 5 minutes.

Figure 17.11: Simulation Model Using the New WorkCell Sub-Simulation Model
Step 4: For the Cell1, make sure the Load Delay Time property is one minute while the UnloadDelay time
property is 0.5 minutes.

Figure 17.12: Setting the Properties of the WorkCell

Step 5: Save and run the model for 24 hours looking at the results. Notice that the statistics of all the pieces
of the work cell are visible under the WORKCELL object type.

Question 2: For a single 24 hour replication, what is the average time in system (in minutes)?
________________________________________________________________________________

Question 3: From an animation standpoint, what did you observe when the parts entered the workcell?
_______________________________________________________________________________

Sub-Modeling: Cellular Manufacturing 293

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 6: Select the WORKCELL Cell1 and choose a different symbol for animation purposes.271

Figure 17.13: Changing the Symbol

Question 4: What did you observe when the parts entered the workcell now?
________________________________________________________________________________

Part 17.4: Adding to the WorkCell Object

The internal view of the composition object (i.e., sub-model) is really the symbol of the object, which when
replaced with a new symbol, the internal workings were removed. In this simple example, showing the internal
workings may make sense but in more complicated sub-models, hiding the details of the sub-model maybe
more beneficial (i.e., just having a plain symbol to represent the object). However, when using a different
symbol, it would be helpful to see some indication of the locations of the entities (i.e., animated queues).

Step 1: Select the WORKCELL object in the [Navigation] panel and select the “Facility” tab. Select all of the
animated queues and then right click to toggle off the Externally Visible option.

Step 2: Go to the Definitions→Externa” section; add four animation queues from the “Animation” tab on
the ribbon for the new work cell object.

· One queue should animate the SrvMachineA. InputBuffer.Contents to represent entities

waiting to be processed while both processing queues of each server
(SrvMachineA.Processing.Contents and SrvMachineB.Processing. Contents)
should be animated.
· Finally, animate the output buffer queue of the SrvMachineB to animate entities waiting to leave the
work cell (SrvMachineB.OutputBuffer.Contents).272
Processing A Processing B

InputBuffer OutputBuffer

Figure 17.14: External View Showing the Four Animated Queues

Step 3: When the WORKCELL’S animation symbol is changed, these four queues will still remain visible.
Navigate back to the original model. The new animated queues will not be visible on objects already

271
In this example, a factory symbol was chosen and it was resized and rotated.
272
Draw the queues from right to left because the first point drawn will be the start/head of the queue.
294 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 declared in the model. You can either add these manually to the current Cell1 object or you can delete it
and add it back to the model making sure to specify the 10 minute paths, changing the symbol, etc.

Question 5: What did you observe when the parts entered the workcell now?
________________________________________________________________________________

Step 4: Run the model observing the animation of the WorkCell when a different symbol has been used.

Currently, the entity will be hidden from animation view when it is on the conveyor as well as waiting for
processing at machine B or any other place that does have an animated queue associated with it. Since the
animation of the internal submodel is not shown, we can reveal some of the behavior by placing labels that
display the number of entities at the various locations. However, it would be helpful to have a representation
of the number of parts currently being processed by the work cell either by machine A or B as well as the
number traveling on the conveyor as well as those waiting in the input buffer of the machine B.

Step 5: SIMIO has the ability to specify user defined storages (queues). Select the WorkCell model in the
[Navigation] panel and navigate to the Definitions tab. Under the Elements section, insert a new STORAGE
element named StorageWorkCellInProcess. We will accept the default ranking rule of “FirstInFirstOut”.

Figure 17.15: Adding a Storage

Step 6: Recall, there are two process steps (i.e., Insert and Remove) that can be used to manipulate user
defined queues.273 Once the part enters processing at SRVMACHINEA, they should be inserted into the
StorageWorkCellInProcess queue to indicate they have gone into processing. Add an Insert step after
the operator has been released in the processing with the Queue State Name property associated with the
new user defined queue WorkCellInprocess.Queue. The Object Type should remain the
“AssociatedObject” because the token is associated with the part.274

Figure 17.16: Using an Insert Step to Add Parts to a User Defined Queue
Step 7: Once the part has finished processing and removed from the machine, delete the part from the user
defined queue using a Remove step as seen in Figure 17.17.

273
These steps do not work on queues defined by SIMIO (which is quite a limitation).
274
The Rank Placement property can be used to specify the placement of the entity in the queue ignoring the overall ranking of the
queue.
Sub-Modeling: Cellular Manufacturing 295

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 17.17: Using a Remove Step to Delete Parts to a User Defined Queue

Step 8: Go to the Definitions→External view section for the WorkCell Object and add an Animation
Queue above the object that animates the StorageWorkCellInProcess.Queue (you may want to
delete to the Processing.Contents queues for machine A and B inserted earlier).

Step 9: Return to the main model and manually insert an animated queue for Cell1, namely Cell1.
StorageWorkCellInProcess.Queue. Recall, none of the external view changes will appear for
objects already inserted.

Step 10: To see that the model is working correctly, insert five additional symbols for the Parts coloring
each one with a different color. Set the Animation→Random Symbol property of the Parts to “True.”
Change the mean interarrival time to be Exponential(2) minutes.

Step 11: Save and run the model observing that the parts are now animated even if they currently are on the
conveyor. The object now acts like a server with an input buffer, processing queue (i.e.,
StorageWorkCellInProcess), and an output buffer.

Question 6: For 24 hour replication, what is the average time in system (in minutes)?
________________________________________________________________________________

Part 17.5: Exposing Resource and Capacity Properties

Currently the operator is hidden inside the WORKCELL sub-facility model and thus, cannot respond to work
schedules. Also, the capacity and processing times of the two machines are hidden to the user of the objects as
well.

Step 1: Return to the WORKCELL model and select the resource ResOperator. All fixed models have built-
in properties of capacity type, capacity, ranking rule, and dynamic selection rules. Therefore, the
ResOperator should be set equal to the WorkCell object’s properties. Select the ResOperator and
specify all of the process logic properties to their associated WorkCell properties as seen in Figure 17.18.
When the user specifies these properties of the WorkCell, the ResOperator properties will be set to the
same.

296 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 17.18: Exposing the Resource Capacity and Selection to the User
Step 2: Each of the fixed model inherited properties by default is hidden. Therefore, you will need to go to
the Definitions→Properties section and expand down the Properties (Inherited) and set the Visible
property to “True” for all the properties in Figure 17.18 as well as change the Category to “Process
Logic”.

Step 3: Create a new reference property for the Processing Time property for each of the two servers
named MachineAProcessingTime and MachineBProcessingTime respectively and the Initial Capacity
property named MachineAInitialCapacity and MachineBInitialCapacity.

Step 4: In the Definitions→Properties section, place all of these new properties into the “Process Logic”
category and set the default units of the processing times to minutes.

Figure 17.19: Machine B Processing Time Property Values

Step 5: Return the main model making sure to set the processing times of machines A and B to be equal to
four and three minutes respectively.

Step 6: Save and run the model observing that nothing has changed.

Step 7: If you were only going to have one cell, then it does not make sense to take the time to create a
sub-model object. However, the benefit of a sub-model object is in reusing the object over and over,
without having to repeat the same process logic and properties. Another benefit is if the logic changes or
additional capabilities are needed, the changes are made in only one place.

Step 8: Insert two new WORKCELL objects named Cell2 and Cell3 by copying Cell1 object to include the
same symbol with the same processing times of four and three minutes but make the capacities equal to
two for the capacities for the two machines. Notice the new animated queue is automatically added.

Sub-Modeling: Cellular Manufacturing 297

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 9: Connect the SOURCE and SINK via five minute time paths to these new workcells and increase the
arrival rate to two per minute.

Step 10: Save and run the model observing the results.

Figure 17.20: Demonstrating the Reuse of the WorkCell Object

Step 11: Next create a DATA SCHEDULE named SchedCell. Change the Standard Day such that there is a
repeating schedule with three hours being “Off Shift” and three hours being “On Shift”. “On shift” should
have a capacity of “4” as seen in Figure 17.21.

Step 12: Select all three work cells and specify the Capacity Type property to follow a WORKSCHEDULE with
the Work Schedule property the SchedCell as seen below.

Figure 17.21: Specifying the WorkCells follow a WorkSchedule

Question 7: What do you notice about the parts in-process in the cell when the shift changes?
________________________________________________________________________________

Question 8: For 24 hour replication, what is the average time in system (in minutes)?
________________________________________________________________________________

Part 17.6: Passing Information between the Model and its Sub-Models

Sometimes the sub-model must be constructed using information from the model, but before the model has
itself been constructed. For example, suppose the WorkCell needs an operator for MachineA, but the
operator is supplied from the model and not known to the sub-model. The sub-model needs to be created
knowing that an operator will be available, but it doesn’t know its exact specification. Here a property
“reference” is useful for the sub-model, expecting that the property instance will be supplied in the model.

Step 1: Using the model from the previous section, select the WorkCell from the [Navigation] panel.
From the Defintions→Lists add a new Object List Property from the Object Reference dropdown named
298 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 ResourceList. Set the Required Value property to False, since the main model will be supplying the
actual list.

Step 2: Click on the SrvMachineA server and use the Secondary Resources section using the Resource for
Processing specification to select a resource randomly from the ResourceList as shown in Figure 17.22.

Figure 17.22: Specifying the Resource

Step 3: Select the MODEL and add two resources named ResOperator1 and ResOperator2. Within the
Definitions tab and the Lists option, add a new list of objects. Call it GResourceList and have it contain
the two operators.

Step 4: By clicking on each cell, specify the GResourceList for the individual cell’s ResoruceList. Doing
this will link the model list to each of the sub-model lists.

Step 5: Run the model to see if it appears to be executing appropriately.

Step 6: Currently, we do not know where the operators are working since they are working at the sub-
model level. At the model level, add an INTEGER STATE VARIABLE to the MODELENTITY called
EStaCellNumber, which will be used to identify the cell.

Step 7: In the State Assignments section of the TIMEPATHS between the SrcParts and the cells, use the
Before Exiting specification to assign the cell number to the EStaCellNumber. Figure 17.23 shows the
assignment on the path to Cell 3. Repeat for the other two cells.275

Figure 17.23: Assigning the Cell Number

Step 8: Add Status Labels under each resource to dispaly which cell the resource is working (refer to
Figure 17.24). Show the status of each of the resources. The following expression is for ResOperator1
which first checks to make sure the RESOURCE has been seized and if it has will display the entities cell
number.
Math.If(ResOperator1.ResourceOwner > 0,
ResOperator1.ResourceOwners.FirstItem.ModelEntity.EStaCellNumber, 0)

275
This is another great place to use a Tokenized process to set the cell number instead of using Before Exiting assignments especially
if the number of work cells increased way beyond three.
Sub-Modeling: Cellular Manufacturing 299

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 17.24: Communicating With the Sub-Model
Question 9: Is the model executing at expected?
________________________________________________________________________________

Part 17.7: Commentary

· The idea behind the sub-model is that you can create your own object class by “composing” a
model and then using it as an object. Once that object class is created/defined, you can use the
object from that class multiple times in the same project or in different projects.

300 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Chapter 18
The Anatomy of Objects: Server
Many simulation languages do not have the ability to modify the behavior of the objects and thus require
another object to perform supporting activities (reneging is an example). Unlike most simulation languages, if
you do not like the way the designers of SIMIO implemented the SERVER object (or any other SIMIO object)
you can modify it or create your own object.

Part 18.1: A Simple Resource Model: Warehouse Pickup

A warehouse store has a pick-up area where customers come for items that are not stored on shelves. The
pick-up area is serviced by an associate who waits on those customers. Because the pick-up area does not
require the full attention of the associate, the associate is also asked to do some specialty stocking. Quite
appropriately, the store wants the customers serviced before doing any stocking.

Exponentially(2) minutes Exponentially(1.5)

Priority One minutes
Customers Serve
Arrive 10 m Customers 10 m

Actual Customers Have Exit

Priority over Stocking One Associate

Stock Items in 10 m
Items 10 m
Arrive Storage
Exponentially(4) minutes Exponentially(2.5)
Priority Zero minutes

Figure 18.1: Warehouse Operations

Step 1: Create a new model.

Step 2: Insert two SOURCES named SrcCustomers and SrcItems and two MODELENTITIES named
EntCustomers and EntItems.

· The SrcCustomers should create EntCustomers while the SrcItems creates EntItems.
· Customers arrive exponentially every two minutes while items arrive exponentially every four
minutes.
· Color the EntCustomers red.
· Set the Initial Priority to zero for EntItems since by default MODELENTITIES have a priority of one.

Step 3: Insert two SERVERS named SrvCustomer and SrvStorage and one SINK named SnkExit.

· Each of the SERVERS should have a capacity of one.

· The Processing Time property for SrvCustomer should be set to Exponential (1.5) minutes and
Exponential (2.5) for SrvStorage.
· Connect the two SOURCES to their appropriate SERVER via 10 meter logical paths and the two SERVERS
to the SINK via ten meter paths as seen in Figure 18.2.
The Anatomy of Objects: Server 301

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 18.2: Model of a Warehouse Pick-up Stations
Step 4: Insert a fixed RESOURCE named ResAssociate between the two SERVERS as seen in Figure 18.2 to
constrain the processing of entities.

· Change the ResAssociate’s Dynamic Selection Rule property to be “Largest Value First” with the
Value Expression left as the default property (i.e., Candidate.Entity.Priority). This should
force the associate to service the customers first if there is a choice.
· Recall the RESOURCE has seven defined states defined by the ResourceState variable. In order to
visually indicate the type of part that is currently being served, add an additional symbol coloring the
round circle red and the outside green for the 8th symbol to represent busy for red parts. Then select
the 1st symbol and color the iris green as seen in Figure 18.3.
Other Busy
State

Red
Figure 18.3: Adding a Second Symbol for Associate
Step 5: From the Definitions→States section, add a new DISCRETE INTEGER STATE variable named
GStaAssociatePicture to be used to specify the symbol associated with the customer type since the
RESOURCE picture cannot be specified directly.

Step 6: Change the Current Symbol Index of the property of the ResAssociate to use the new state variable
as shown in Figure 18.4. If the state is busy and the entity is a customer then display the 7th symbol
otherwise show the default state symbol.

Figure 18.4: Animation Symbol Index

Step 7: From the Processes tab, create two new processes named SeizeAssociate and ReleaseAssociate.
The ReleaseAssociate process should release the associate while the SeizeAssociate should
seize the associate as well assign the GStaAssociatePicture the value of the
ModelEntity.Priority as seen in Figure 18.5.

302 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 18.5: Processes for Seizing and Releasing the Associate
Step 8: For both the SrvCustomer and SrvStorage, specify the Before_Processing add-on process trigger
to be the SeizeAssociate process and the After_Processing trigger to be the ReleaseAssociate in similar
fashion.276

Step 9: Save and run the model.

Question 1: When the associate finishes with a customer, does he always attend to another customer if
they are waiting?
________________________________________________________________________________

Question 2: Why does this happen, since the Dynamic Selection criteria of the ResAssociate was specified
correctly?
________________________________________________________________________________

Step 10: Select the ResAssociate and from the Attached Animation section, insert the Allocation Queue
drawing it from left to right somewhere in the model. The allocation queue represents all the requests for
the ResAssociate and will be ordered by priority. Save and rerun the model.

Question 3: What do you notice about the arrival of entities with regard to the queue?
________________________________________________________________________________

Allocation
Queue

Figure 18.6: Causing the Allocation Queue to Become Visible

Part 18.2: Taking an Object Apart to Figure out how it Works

For some reason, a customer waiting at the SrvCustomer does not register a request for service when the
associate finishes with another customer. At this point, the problem may not be evident. Unlike other

276
You can select both of them at the same time and specify the add-on process triggers.
The Anatomy of Objects: Server 303

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 simulation languages, SIMIO objects can be examined by taking them apart, and possibly modifying the
behavior. Existing objects can examined as well as modified in SIMIO via “subclassing.”

To try to understand why the queue has not registered the other customers, the SERVER will be sub-classed to
understand its inner workings. Subclassing provides the ability to extend and modify an existing object.

Step 1: To subclass any of the standard SIMIO objects, right mouse click on the object in the standard
library and choose subclass in the menu option as seen in Figure 18.7.

Figure 18.7: Subclassing the Server

Step 2: After subclassing the SERVER, select the MYSERVER object in the [Navigation] panel. Now all the
characteristics of the SERVER become visible. Look at the Definitions→Element’s section. The SERVER
object defines ten different elements: one COSTCENTER, one FAILURE, three STATIONS, one
TASKSEQUENCE, and four TIMER events.

· The COSTCENTER element is used to define the parent activity-based costing for the object.
· The FAILURE element is used to define a failure mode. Failure downtime occurrences are started and
ended using the Fail and Repair steps.
· The STATION element is used to define a capacity constrained location within an object where one or
more visiting entity objects can reside. Each STATION defines an EntryQueue and Contents
Queue. The Contents Queue contains the entities which have seized capacity of the station and are
being processed while the EntryQueue will hold the entities waiting for capacity.
· The TASKSEQUENCE element is used to reference the set of tasks that compose the processing time
when the Process Type is Task Sequence.
· The TIMER element fires a stream of events according to a specified IntervalType (i.e., calendar time,
processing count, event count, or processing time). These timers determine when failures occur if a
failure has been defined.

When entities arrive at the input node they are automatically transferred to the “InputBuffer” Station if it has
capacity. From the “InputBuffer”, the entities are transferred to the “Processing” station and once they have
finished processing they will be transferred to the “OutputBuffer” station as seen in Figure 18.8.

Server
BasicNode TransferNode

InputNode OutputNode
InputBuffer Processing OutputBuffer
Station Station Station

Figure 18.8: Showing the Entity Flow through the Server Object
Step 3: Next, select the “Processes” tab to look at the behaviors. The SERVER object defines 13 processes
that govern how the SERVER reacts to failures, capacity changes, entity arrivals and exits, etc. Some of the
304 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 processes respond directly to events (e.g., entities entering the Processing station, entities entering the
Inputbuffer, etc.). The two processes of interest (see Figure 18.9 and Figure 18.10) are: the OnEntered
InputBuffer which is executed when entities enter the InputBuffer station from the input node if the
input buffer has capacity, and the OnEnteredProcessing which is executed when entities enter the
processing station from either the input buffer or the input node.

Figure 18.9: OnEntered Processes for InputBuffer Station

Figure 18.10: OnEntered Processes for Processing Station

The Anatomy of Objects: Server 305

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 When entities arrive at the SERVER they are transferred to the InputBuffer which automatically executes the
OnEnteredInputBuffer process which has the following steps. Examination of these steps will yield
some insight into our problem.

· Perform any On Entering State Assignments

· Seize and Release On Entering Secondary Resources
· Execute the “Entered” Add-on process trigger. The Execute process step will run the specified
process before proceeding.
· Delay for the transfer time if specified.
· End the transfer of the entity into the InputBuffer STATION.
· Seize one unit of capacity of the SERVER.
· Seize and Release processing secondary Resources.
· Once the entity is able to seize capacity of the SERVER and other secondary resources it will then
execute the “Before_Processing” Add-on process trigger which in the pick-up station model will
seize the associate.
· Once it has finished executing the Before_Processing add-on trigger, the entity is transferred to the
Processing station where the other process takes over.

Once entities are transferred to the processing station the OnEnteredProcessing is invoked where the
actual processing is performed.

· If the input buffer capacity is zero, it goes through the same process as the OnEnteredInput
Buffer (i.e., the false branch of the Decide step).
· End the transfer of the entity into the Processing STATION.
· Execute the “Processing” Add-on process trigger.
· Depending on “Process Type” either start processing tasks or Delay the entity based on the
processing time specified.
· Once the entity has been delayed or the tasks have finished, it will then execute the
“After_Processing” Add-on process trigger which in the pickup station model releases the associate.
· Seize and release any Secondary Resources.
· The entity will either exit the SERVER or be transferred to the OutputBuffer station if the capacity of
the output is greater than zero.

The first clue to understanding our problem is that the Before_Processing add-on process trigger is not
executed until the entity is able to seize capacity of the SERVER.

Step 4: Next examine the OutputBuffer Entered and Exited processes as seen in Figure 18.11.

Figure 18.11: OnEntered and OnExited of the Output Buffer Station.

306 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 The OnEnteredOutputBuffer process will be invoked when the entity is transferred to the
OutputBuffer STATION and goes through the following steps.

· Ends the transfer into the output buffer station.

· After ending the transfer, the entity releases capacity of the SERVER at this point.
· It then performs the before exiting state assignments.
· The entities are then transferred to the output node which invokes the OnExitedOutputBuffer
process which executes the “Exited” add-on process trigger.

Step 5: Now it can be seen that the After_Processing add-on process trigger which releases the associate
happens before the SERVER capacity is released. Therefore, the customer waiting in the input buffer cannot
request the associate before the associate evaluates the current requests since it first has to seize the server
before the SeizeAssociate process is executed.

Step 6: To fix the problem, set the capacity of the two SERVERS to “Infinity” which will eliminate the
SERVER constraints. Therefore, every entity entering the two servers will immediately execute the
After_Processing add-on process trigger and request the associate.

Step 7: Save and run the new model.

Question 4: After observing the model for a while, does the associate always service customers first?
________________________________________________________________________________

Step 8: Setting the capacities to “Infinity” eliminates the problem. However there are situations where the
SERVER capacity should not be infinite because available space in the server may also limit the number of
entities being served. For instance, there may be more than one associate but no room to serve more
customers. Notice from Figure 18.11, the Exited add-on process trigger happens after the Release step
and can be used to release the associate.

· Change the capacity of the SERVERS back to one.

· Remove the After_Processing add-on process trigger from each SEVER by setting the value to null,
removing it manually, or right clicking and then select Reset.
· Choose the ReleaseAssociate as the Exited add-on process trigger for both SERVERS.

Step 9: Save and run the new model. You may want to increase the item arrival rate to one or two per
minute.

Question 5: After observing the model for a while, does the associate always service customers first?
________________________________________________________________________________

In other instances, having the correct order may have resulted in the same phenomena owing to zero time
events which can be checked by looking at the “Model Trace”. Recall one can reorder the zero time events by
placing a Delay step right before a step you want to force to the end of the zero time by delaying it for zero
time using Math.Epsilon as seen in Figure 18.12. When SIMIO encounters a delay of Math.Epsilon it
places the steps after at the end of the current event time.

Figure 18.12: Adding the Zero Time Delay

The Anatomy of Objects: Server 307

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Part 18.3: SIMIO Objects and Class Heirarchy

Objects are created (instantiated) from classes. The act of clicking on a SIMIO object in the [Standard
Library] panel and dropping it into a position on the modeling canvas causes an object or model of that type to
be created (instantiated). The object will possess all the characteristics defined for that class. Once on the
modeling canvas, you can change (even delete) its “Default” characteristics such as initial property values,
processing time, capacity, add-on processes, and so forth. You can also add your own properties, events,
states, and elements. Each SIMIO object will have its own collection of processes, elements, properties,
states, and events. Objects can be saved in a library and used in other objects. SIMIO makes no distinction
between a model and an object.

Figure 18.13 shows the base object classes in SIMIO as well as the [Standard Library] model classes. New
classes can be created from base classes and new classes can be obtain from the [Standard Library] (or your
library) by “sub-classing”.

Intelligent
Object

Agent Link Node Fixed

Connector, Path,
BasicNode,
Entity TimePath, Processor
TransferNode
Conveyor

Source, Sink, Server,

Transporter Workstation, Combiner,
Separator, Resource

Vehicle, Worker

Figure 18.13: SIMIO Class Hierarchy

The “origin” object class is the INTELLIGENT OBJECT (IO) class which gives objects the ability to be seized
and released as well as respond to schedules. The AGENT class inherits the properties and characteristics of IO
class but adds additional characteristics and behaviors (i.e., objects that can be dynamically created and
destroyed as well as move around in continuous and discrete space (grid)). . The LINK adds the ability (i.e.,
properties and behaviors) for ENTITIES to enter, exit and move along. ENTITIES are specialized agents that
have the ability to move on links (paths, conveyors, and connectors) as well as move in and out of FIXED
objects (combiners, servers, workstations). FIXED objects represent intelligent objects that are fixed and don’t
move in the space. The TRANSPORTER is a more specialized entity that has the additional capability of being
able to pick up and drop off entities. They can do this by moving in free space or along links. At present
models cannot be created directly from the INTELLIGENT OBJECT or from the ENTITY classes.

New object classes in SIMIO may be defined by extending or modifying the current SIMIO objects. In object
oriented terms this method of extension is referred to as specialization or “subclassing.” When a model is
subclassed, it inherits all of the characteristics (i.e., processes, elements, properties, states, and events) but has
the capability to modify those inheritable features as well as add new ones.

308 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 The object classes in the SIMIO [Standard Library] should be viewed as model classes which have been
specialized from other classes. Part of their characteristics are derived from the base classes, but some
characteristics are added or changed. For example, a WORKER object is a specialization of an entity that can act
as a RESOURCE and a VEHICLE. WORKERS and VEHICLES are specialized TRANSPORTERS. Several properties (e.g.,
Initial Node, Idle Condition, etc.) can be modified and added as well as adding new process logic to force the
entity to behave similar to resources and vehicles. The LINK, NODE and SERVER objects have the same
properties as the FIXED class, but add more functionality. Likewise the PATH, CONNECTOR, and CONVEYOR
objects are more specialized versions of the LINK class. The SERVER class is a fairly general object used to
constrain and delay entities in the system. The PROCESSOR class is a specialization of the FIXED class

The ability to extend and add objects is an important feature of SIMIO that needs to be studied carefully. Thus
if you are only going to need the sewing machine class for one instance of one model, you may not want to
take the time to create an object class. However, if you’re going to use this class in multiple places, then
having a reusable object will be important and its utility will be demonstrated over the next few chapters.

SIMIO uses a three tier object hierarchy composed of the “object definition”, the “object instance”, and the
“object realization”. The object definition is simply the specification of the object class through its processes,
states, elements, properties, and events. The definition may be determined by the SIMIO developers for the
base classes and [Standard Library] objects, however a modeler may create their own object definitions (as we
will see in the next several chapters). The object definition is shared by all instances from that class. While
all objects from the same class share the same definition, each object instance has its own characteristics such
as their own add-on process, state variable value, capacity, etc. The object instance may have its
characteristics changed during the simulation.

Object realizations hold the state values for the instances and these occupy a fixed location in the model.
However ENTITIES, VEHICLES, and WORKERS, are dynamic and are created from the instance definition during
the execution of the simulation. So SIMIO creates static representation for that part of these object definitions
that are common, but then creates a dynamic representation for the instances that move throughout the
simulation. You can see this in the model with a vehicle, whose object realization shows on the modeling
canvas, while the individual vehicles move around the model.

The Anatomy of Objects: Server 309

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Chapter 19
Building New Objects via Sub-Classing: A Delay Object
In the prior chapter we demonstrated how specializing/subclassing existing objects allows the characteristics
and behaviors of an object to be exposed. The subclassed object may have characteristics and behaviors added
or existing ones to be modified. To illustrate we will create a new object through subclassing an existing
objects and change its character. In particular, we address the need for a delay node. Sometime entities need
to be delayed for a certain amount of time before continuing to their next destination. The SERVER object can
be used to do this, but it is really more complicated than is needed to perform a simple delay. We will extend
the TRANSFERNODE by adding the ability to delay entities that pass through it.

Part 19.1: Sub-Classing the TRANSFERNODE to Create a DELAYNODE

The TRANSFERNODE is an object that is used to transfer entities out of standard objects (e.g., SERVER, SOURCE,
etc.) or as transfer locations in a network of links. It provides an excellent choice to delay an entity before
continuing to its next destination. The logic to ride on transporters, continue by sequence, and take the
shortest path is already built into the object and these properties and behaviors do not need to be reinvented in
a new object. Therefore, the TRANSFERNODE will be specialized/sub-classed. This approach illustrates how an
existing object is extended and modified. Objects created from existing objects are referred to as a “derived”
objects.

Step 1: First create a new project and model in SIMIO.

Step 2: To subclass or “derive” any of the standard objects277, right mouse click on the object and choose
subclass in the menu option as seen in Figure 19.1.

Figure 19.1: Sub-Classing the TRANSFERNODE

Step 3: A “derived” version of the TRANSFERNODE class is placed in the [Navigation] panel as
MYTRANSFERNODE. Right click the new class and select the Model Properties option to display all of the
properties.

· Change the Model Name property to DELAYTRANSFERNODE.

· Change the Icon property to any appropriate icon (i.e., a png or bitmap file).278 The icon is the picture
that is seen in the project window that is used to select the object. For example, the Delay.png file
seen in Figure 19.2 will be used in the chapter.

277
You can also subclass any user created object as well.
278
Icons are typically 32 pixels wide by 32 pixels high. Many image editing software can create these icons (e.g., Photoshop, Paint or
freeware GIMP program. (http://www.gimp.org)).
310 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 19.2: Icon for the DELAYTRANSFERNODE

· Under the “General” category, change the color of the new DELAYTRANSFERNODE to something like
maroon.279
· Select the new DELAYTRANSFERNODE in the [Navigation] panel and let’s take a moment to look at the
processes and characteristics that were inherited. Navigate to the Processes tab and see the processes
(i.e., behaviors) in Table 19.1 that have been inherited from the parent TRANSFERNODE class.

Table 19.1: Processes Inherited from the Parent TransferNode

Process Event Process Description

This process runs after the leading edge of an object (ENTITY,
OnEntered
TRANSPORTER, or AGENT) has entered the node.
This process runs after an ENTITY has entered the node from
OnEnteredFromAssociatedObject
another object (i.e., dropped off by a VEHICLE).
This process runs after the leading edge of an object (ENTITY,
OnEnteredParking
TRANSPORTER, or AGENT) has entered the parking station.
This process runs after an ENTITY has entered the node and is
OnEnteredToAssociatedObject heading to an associated object. An object uses the node as its
input node (e.g., entering the SERVER object.).
This process runs after the trailing edge of an object has
OnExited
exited the node.
OnRunEnding This process runs when the simulation ends.
OnRunInitialized This process runs when the simulation starts initially.
This process runs when the ENTITY or TRANSPORTER enters
OnEnteredParking
the parking station which only ends the transfer.
A process used to handle the logic of routing objects out of the
RoutingOutLogic
node (i.e., ride on a TRANSPORTER, by sequence, etc.)
If an object fails to transfer to an outbound link it will destroy
TransferFailureLogic the object or park it if it fails (i.e., Transporters that cannot be
destroyed).
· Select the Definitions tab to access the characteristics of the node. Elements represent objects in a process
that change state over time (i.e., statistics, timers, monitors, failures, etc.). There are two elements defined:
a parking STATION and a ROUTING GROUP.
· The PARKINGSTATION station is used to house entities, operators, and transporters parked at the node.
The station element will be discussed in more detail in the next chapter.
· The ROUTINGGROUP element is used with a Route step to route an entity object to a destination
selected from a list of candidate nodes. The TRANSFERNODE object uses a RoutingGroup element in
its internal logic to determine where the entities will travel when they leave the node.
· The new sub-classed node inherits several properties and states as well as four events from the parent
TRANSFERNODE as described in Table 19.2.280

279
You cannot change the picture (i.e., diamond) because nodes are special objects that take up zero space in the actual system.
280
To access Inherited items, click on the to make the items visible.
Building New Objects via Subclassing: A Delay Object 311

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Table 19.2: Events Inherited from the Parent TransferNode
Event Event Description
CapacityChanged The node’s capacity has changed.
Entered An object’s leading edge has entered the node.
Exited An object’s trailing edge has left the node.
RiderWaiting An ENTITY is waiting for transport at this node.
·

· Two string lists (i.e., TransferNodeDestinationType and TransferNodeTallyConditionType) have

been defined which are used in property dropdowns. The TransferNodeDestinationType list is used
to specify the property of how to route the entity to its next destination node (i.e., Continue, Specific,
BySequence, and SelectFromList) while the TransferNodeTallyConditionType is the condition type
(NoCondition, IsEntity, IsTransporter, CustomCondition) used to compute Tally statistics at this node.

Part 19.2: Modifying Processes and Adding a Properties for the New Node

The new delay node sub-classed from the standard TRANSFERNODE will operate just like the original node since
nothing was added or modified. The node needs to delay the objects that enter the node either individually or
via a transporter.

Step 1: Under the Definitions tab in the delay node, add an Expression property281 named DelayTime that
can be used by a user to specify the delay time for the entity. Specify the Default Value property to be zero
while the Unit Type should be “Time” with a Default Units of “minutes” as seen in Figure 19.3. Place the
new property underneath the “Process Logic” category.

Figure 19.3: Inserting a Delay Expression Property

Every time, an entity enters this transfer node either on its own or via an associated object, it should be
delayed by the delay time expression property. Therefore, the OnEnteredFromAssociatedObject282
and the OnEntered processes will need to be modified to reflect this consideration. Currently the processes
are not editable, which can be seen by selecting the OnEntered process and noticing the steps and properties
are grayed out. After selecting the process,283 click the Override button ( )which will allow changes to be

281
The Expression Property allows the user to specify any mathematical expression similar to a Processing Time of the SERVER.
282
For example, the object is entering via transporter.
283
Select the Processes tab to access all the processes.
312 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 made. Once the process has been overridden the Override symbol is placed by the process name to indicate it
has been modified. Note, the Restore button ( ) can be used to return the process back to its original state.

Figure 19.4: Overriding and Restoring Inherited Processes

The OnEntered process is fairly complex with the following basic steps.

· On Entering Tally statistics to be collected if specified.

· Fires the “Entered” Event.
· Then Executes the “EnteredAddOnProcess” that a user may have specified.
· Checks to see if node is “Bound to an External Node” and if so transfers it to the specified parent
external node.
· The Visit Node step is used to invoke the entity’s OnVisitingNode process to indicate the
entity has arrived at the destination node.
· If this entity is a transporter then transfer it to the outbound link since it does not go through the same
logic and if there is no outbound link, the transporter is sent to free space or processes a transfer error.
· If the entity is not a transporter, then it executes the RoutingOutLogic process to transfer the
entity out of the node with the correct logic (i.e., uses the “Route” step in conjunction with the
ROUTINGGROUP element). If not riding on transporter in routing out, then the transfer to the outbound
link occurs and the prior logic applies.

Question 1: Where should the delay step be inserted into the process?
________________________________________________________________________________

Step 2: In some cases, users may want change the delay time based on some condition when they enter the
node. Therefore, insert a Delay step after the Entered Add On Process is executed to allow the
processing time to be potentially set in the add on process before the delay occurs. The Delay Time
property should be set to the DelayTime reference set in the previous step as seen in Figure 19.5.284 Under
the General Section, the Name property can be specified to be “DelayTime” for a better description.

Figure 19.5: Modifying the OnEntered Process to Include the Delay

284
Right click on the property label to specify a referenced property.
Building New Objects via Subclassing: A Delay Object 313

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 3: Repeat the procedure for the OnEnteredFromAssociatedObject process by adding the
Delay step in the same location. Once the process has been overridden, just copy the delay step and paste
the step into the appropriate place in this process.

Step 4: We may want to add the delay to the OnEnteredToAssociatedObject if you were to use as
an external node for another object. Just override the process and copy the Delay step between the
execution of the Entered_AddOnProcess and the decision if the node is bound to an external node.

Part 19.3: Creating a Model to Test the New DELAYTRANSFERNODE

Now that the new delay transfer node has been created, it can now be used in any model where the original
TRANSFERNODE is used with the added benefit of delaying entities for some time as they enter the node. Figure
19.6 shows a network where we will test our new DELAYTRANSFERNODE.

5 minutes 10 minutes
(Every 5 minutes)
DelayTransfernode
Parts Create Exact Exit
Arrive Connector Copy

5 minutes 10 minutes

Normal Transfernode

Figure 19.6: Building an Example to Test our DelayTransferNode

Step 1: Go back to the original Model by selecting it in the Navigation panel.

Figure 19.7: Navigating Back to the Original Fixed Model

The model as seen in Figure 19.8 will create an entity and then duplicate it sending the original through a
DELAYTRANSFERNODE and the duplicate through a standard TRANSFERNODE. The time in the system will be
used to illustrate the delaying behavior.

· Insert a new MODELENTITY named EntParts.

· Insert a SOURCE named SrcParts that creates EntParts.
· Add a SEPARATOR named SepCopy which can be used or to create copies of entities or to un-batch
entities that have been combined/batched using a Combiner.
· Add a new DELAYTRANSFERNODE named TDelay which can be accessed via [Project Library] panel.
· Add a normal TRANSFERNODE named TNormal
· Add a sink named SnkExit.
· Connect the SrcParts to the SEPARATOR by a connector.
· Connect the parent output to your new TDelay and the member output to the TNormal node via
TIMEPATHS that each takes five minutes.
314 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 · Connect both TNormal and TDelay to the SINK via ten minute time paths.

Figure 19.8: Model Used to Demonstrate the New Delay TransferNode

Step 2: Configure all the fixed objects with the following parameters.

· SOURCE: Parts should arrive at a constant interarrival time of five minutes.

· SEPARATOR: Separation Mode property should be specified to “Make Copies” and the Copy Quantity
property should be one, as seen in Figure 19.9. This will make one copy of the current entity. All
properties of the entity will be copied. However, the entity’s creation time is the current time and not
necessarily the same time as the original entity. The original entity will flow out of the ParentOutput
node while the copy will flow out of the MemberOutput node.

Figure 19.9: Separator Properties

Step 3: A new characteristic needs to be added to the EntParts MODEL ENTITY to track the current time at
certain points in the system. Select the ModelEntity in the [Navigation] panel and add a new Discrete
Real State variable named EStaTimeClock underneath the Definitions tab. A real state variable is needed
since the value will change throughout the run of the system. Make the Unit Type Time and the Units
Minutes.

Step 4: Let’s add a status label to the ENTITY that displays its associated EStaTimeClock state variable.
Select the EntParts object in your model and then choose the Status Label under the Symbols→Attached
Animation section.285 Draw a slender status label near the entity with the expression equal to
String.Format("{0:0.###}",EStaTimeClock)286 seen in Figure 19.10.

285 We could put the expression in the Dynamic Label Text under the Animation section of the entity as well but this is more
visible.
286
This will format the time clock to have three digits of precision.
Building New Objects via Subclassing: A Delay Object 315

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 19.10: Creating a Status Label to Track the Part’s EStaTimeClock State

Step 5: Since the status label is attached to the MODELENTITY it will travel with it. Therefore, rotate (via
Ctrl→Mouse), size and move the label so it sits right on top of the entity as seen in the Figure 19.11
below.

Figure 19.11: Attaching the Status Label to the Part Entity

Step 6: Save and run the model to make sure it behaves as predicted.

Question 2: Does the part get duplicated at the separator correctly?

________________________________________________________________________________

Question 3: Does the status label travel with the EntParts and what is the value of all the labels?
________________________________________________________________________________

Step 7: The value of the EStaTimeClock state variable has not been updated at this point. Navigate to the
“Processes” tab and create a new process named EStaTimeClock. Insert an Assign step that sets the
ModelEntity.EStaTimeClock to a new value of 60*Run.TimeNow287 which set the variable to the
current simulation time in minutes.

Figure 19.12: Creating a Generic Process to Set the Model Entity’s StaTimeClock

Step 8: When the Parts exit the SEPARATOR either through the Parent or Member output, the Parts
EStaTimeClock variable needs to be set to the current time. Select the ParentOutput@SepCopy transfer
node and specify the SetTimeClock process for the Entered add-on process trigger which will assign the

287
Run.TimeNow is a function that returns the current time of the simulation in hours.
316 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 current simulation time to the EntPart’s EStaTimeClock. Perform the same process for the
MemberOutput@SepCopy Entered add-on process trigger.

Step 9: Repeat the process for the Entered and Exited add-on process triggers for both the TNormal and
TDelay transfer nodes.

Step 10: Save and run the model. Observe the time stamps as the entities leave the separator as well as the
times to enter and exit the transfer nodes. You would think that the time stamps should be the same since
the delay time is zero and whole numbers are used for the arrival and time paths. You may want to set a
breakpoint on one of the nodes and then single step until the entity leaves the transfer node.

Question 4: Are the entering and exiting times different for the parts and if so why?
________________________________________________________________________________

It seems there is a small amount time being delayed in one of the transfer nodes as compared to one
another. They enter the nodes at the same time (i.e., at time five in the left side of Figure 19.13) but
appear to leave at slightly different times as seen in the right side of Figure 19.13. However, transfer
nodes take up zero space and therefore should not have caused any delay (i.e., discrepancy). The time an
entity enters the node should be equal to time it left.

Figure 19.13: Showing Discrepancies Between Exited and Entered Times

The discrepancy deals with when the Exited process event is fired. This event will not fire until the
trailing edge of the entity has left the node, which depends on the length of the entity and how fast the
entity is transferred onto the link (i.e., function of the entity’s speed and the speed of the path its
leaving and the one its entering). The Entered process event is fired once the leading edge has entered
the node. Therefore, when setting time stamps the add-on process trigger selected should be based on
the leading edge of an ENTITY. Figure 19.14 demonstrates that both entities enter the node at the same
time. However, the smaller entity will exit the node first as shown.

Building New Objects via Subclassing: A Delay Object 317

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Entered – Leading Edge Exited – Trailing Edge

TransferNode TransferNode
Speed and Size of Entities as
well as the Link speed Effects
Time to Exit Node

TransferNode TransferNode

Figure 19.14: Demonstrating Entering and Exiting from Nodes

Step 11: Reset (remove) the Exited add-on processes of both transfer nodes (i.e., TDelay and TNormal) and
reset the Entered the two outputs of the SepCopy to remove the add-on process trigger.288 Specify the
SetTimeClock process for the Entered add-on process for each of the four time paths leaving the two
transfer nodes and the SEPARATOR.

Step 12: Save and run the model to demonstrate that the entities are correctly synced.

Question 5: Are the time stamps whole numbers and synced?

________________________________________________________________________________

Step 13: Next, specify a delay time of four minutes for the TDelay transfer node and rerun the model. The
part entities are now being delayed four minutes before proceeding into the rest of the network.

Step 14: Change the arrival rate to be every two minutes and observe what happens.

Question 6: What happens to multiple entities that arrive at the delay node?
________________________________________________________________________________

Part 19.4: Commentary

· Derived objects are created by subclassing in SIMIO. Derived objects add or change
characteristics and behaviors of existing objects. In this chapter, delay features were added to the
transfer node and the processing modified accordingly.

288
Recall to reset a process trigger, right click the process and select the “Reset” item.
318 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Chapter 20 Creating New Objects
Traditionally an object is created (instantiated) from an object class. The class describes an object’s
composition in terms of the attributes an object has and the methods (procedures) it performs. In SIMIO the
model itself is an object (from the FIXED CLASS). When SIMIO is initiated, a MODEL and a MODELENTITY is
automatically created within a new PROJECT LIBRARY. The MODELENTITY defines the DEFAULTENTITY as part
of the project and instances of this run-time class can be created throughout the simulation run. The MODEL is
formed by combining instances/objects from the libraries available, usually including the [Standard Library].
It is not possible in SIMIO to create completely independent simulation objects as one might do in a
programming language. Models must be formed from either the base classes or by “subclassing” an existing
object class. The five base classes are: FIXED CLASS, ENTITY CLASS, TRANSPORTER CLASS, NODE CLASS,
and LINK CLASS. There is a PROCESSOR class that is derived from the FIXED CLASS to help in creating
processing objects.

When developing new objects, the user must adopt an “architect’s view” in addition to a “user’s view” since
few object class are developed to create only a single object. Instead, new object classes, like a SERVER or
TRANSFER NODE in SIMIO, is expected to be used many times in one or more simulation projects. Therefore
the purpose or scope of an object’s use becomes central to its design. SIMIO object classes have been
developed for many purposes and with diverse users in mind.

Consider the SERVER object class which is basically designed as a single queue (input buffer) with fixed or
scheduled multiple identical servers (at processing). However, it allows for many specifications including
static and dynamic ranking, transfer-in time, flexible processing time with the option of tasks, buffer
capacities, material handling, reliability of servers, activity-based accounting, secondary resources, state
assignments, etc. In other words, this object class is designed for a wide variety of uses. While state
assignments, the acquisition of additional resources, and many of the other features could be user-implemented
in SIMIO processes, they are made “convenient” specifications for users not wanting to write processes,
making the object class more accessible to casual or learning users. As you review the design of the SERVER
you can begin to understand the architectural “paradox” – namely, you want your objects to be “powerful”
with lots of features for widespread application, and yet you want them to be “simple” so they can be quickly
understood and used. It’s not an easy compromise. Generally the SIMIO objects tend to favor the powerful
side of the paradox while in our development, we will want to explore the simple side.

Typically, object classes with lots of features like those in SIMIO are called “heavyweight” because each
object requires a lot of information and has numerous procedures.289 Lightweight objects may have limited
use but are easy to understand and use, and, if designed properly, can become easy to extend through object
inheritance and/or composition. This chapter will focus on the creation of a lightweight specialized object,
with limited purpose (as compared to SIMIO objects). But in this creation, significant reference is made to the
SIMIO objects as guides for development and the new object class will become a part of our custom library.
Also we use the words “object” and “model” interchangeably which SIMIO also uses.

We begin with a simple delay object and then embellish this object so that it provides a simple queuing
capability. It will eventually be able to create multi-server, single queue objects.

Part 20.1: Creating a Simple Delay Object

In the previous chapter, a new delay transfer object was created by subclassing the standard TRANSFERNODE to
delay entities that enter the node. However, the entities will queue up at the transfer node right on top of one
another. Also, there is no opportunity to add a different symbol (i.e., an oven, etc.) to represent the delay

289
The Vehicle/Worker is clearly the most complicated of the SIMIO standard object classes.
Creating New Objects 319

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 location since the transfer node takes up zero space. Initially, a new fixed object (i.e., a simple delay object)
will be built that only delays an entity in its path, similar to any of the standard objects. Of course, such a
delay can be accomplished by an infinite capacity SERVER or even just a TIMEPATH, but creating this simple
delay object will be instructive both in its creation and as a base for extension.

Step 1: Use the previous chapter model and save it as Chapter20-1.spfx.

Step 2: Insert a “New Model” icon by selecting the “Fixed Class” option to produce a new type of object
class and name it DelayObject. In the “Advanced Options” properties of this object class, be sure that its
Resource Object property is “False” since we don’t plan to use these objects as resources.290

Step 3: Within this new object class, add an expression property named DelayTime. Its Default Value is
Random.Pert(1,2,4,1).291 The Unit Type should be Time and the Default Units would be Minutes.
Finally specify “Process Logic” as the Category Name.

Step 4: Objects that contain entities (SOURCES, SERVERS, COMBINERS, SEPARATORS, WORKSTATIONS, etc.)
utilize STATION elements to house these entities. The STATION element may be used to define a capacity
constrained location within an object, where one or more visiting entity objects can reside. From the
Definitions→Element, insert a new STATION named StationDelay. Its Initial Capacity should be Infinity
because we don’t want to restrict the number of entities using the station.

The station element has two queues associated with each station: Contents and EntryQueue. The
EntryQueue is the queue where entities wait until there is available capacity to allow the entities to
move into the Contents queue. If you specify that the input buffer is ten for a server and the eleventh
entity arrives to the server, the entity will be placed into the EntryQueue and not be allowed to enter the
server (i.e., it will stay at the input node of the server).

Step 5: The delay STATION element provides a “place” for entities to reside while they are being delayed.
However, entities need a way to enter and exit the DelayObject. The external view is used to create the
“visual” representation as well as provide enter and exit nodes to the object. Go to the “External” section
underneath the Definitions tab to create the visual interface.

· First, we need to define the “symbol” that will represent objects from this class. You can use a
symbol from the existing SIMIO symbol library, create your own using the Drawing tab, download a
symbol from the Trimble 3D Warehouse, or import from several cad programs (e.g., SketchUp ©
2015 from Trimble Navigation Limited). In this case, we created the symbol using the drawing shown
in Figure 20.1.
· From the Drawing section, insert an Ellipse centering it on the grid. In the Object section, change the
Object Height to 0.3 and the Line Width to 0.02. From the 3D view, color the top grey and the side
and line black.
· Next insert a half diamond to represent the delay. Using the Polygon tool just define the top, the right
and bottom points. Place the shape so it is positioned halfway on the ellipse. Switch into 3-D view,
select the shape and while holding the Shift key down, raise the label to its just barely above the
ellipse. Switch back to the 2-D view, to center in the in Figure 20.1.

290 Most SIMIO objects have capacity.

291 Since the expression editor is not active, type the expression directly into the Default Value. We will tend to use the random number
stream specification throughout to aid comparisons.
320 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 StationDelay.Contents External Transfer
queue Node

External Basic Node

Figure 20.1: Delay Symbol

Step 6: Add the Input and Output nodes which provide the mechanism for objects to arrive and leave the
delay object. Click the External Node from the Drawing→Transfers section to add two external nodes and
position them to edges of the object as seen in Figure 20.1. External nodes have specifications explained in
Table 20.1 that need to be defined.

Table 20.1: External Node Specifications

Specification Description
The type of node this external node should be (BASICNODE, TRANSFERNODE,
Node Class DELAYTRANSFERNODE, etc.). Transfer nodes are typically used as exit/output nodes since
Name they contain the logic to route entities to the next destination while basic nodes are used as
entrance/input nodes.
Location type in the object that an entity transfers into using this external node (e.g.,
Input
Station, Node, or Container). If you have added objects to the facility window and you
Location
want to transfer the entities to one of the nodes in the fixed model of the new object,
Type
select “Node.”
Which station the entity will transfer into from this node if “Station” is the Input Location
Station Name
Type.
Which node in the facility of the object should entities be transferred into from this
Node Name
external node if “Node” is the Input Location Type.
Container Which container in the facility of the object should entities be transferred into from this
Name external node if the Input Location Type is “Container.”
· The Input node should be a BASICNODE as specified in Figure 20.2. When entities enter the node, we
want them to be transferred automatically to the STATION StationDelay by SIMIO.
· The Output node will be a TRANSFERNODE to allow entities to be routed and/or ride on transporter.

Figure 20.2: Specification of the External Nodes

Step 7: As entities enter the delay object and are delayed, the entities need to be visually seen in the delay
(like the Server.Processing.Contents). Insert the queue animation above the symbol by
selecting the Queue under the “Animation” tab as shown in Figure 20.1. The QueueState should be the
Contents Queue of the StationDelay (i.e., StationDelay.Contents).

Creating New Objects 321

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 8: Since we have defined the default external view, the logic to process the entities entering the
objects needs to be specified. Stations automatically respond to three events as described in Table 20.2.

Table 20.2: Events at a Station

Event Description
Entered Fired when entities leading edge enters the station
Exited Fired when entities exit the station
CapacityChanged Fired when capacity of the station changes

Step 9: Create a new process named OnEnteredDelay which will be executed every time an entity
enters the STATION according to the Triggering Event Name StationDelay.Entered as seen
in Figure 20.3.

Figure 20.3: Specifying the Entered Triggering Event of the Delay Process

Step 10: The first step for the entity token is to end the transfer into the station using an End
Transfer step which indicates the ENTITY has left its present location and finished the transfer
into this location. This step also fires the Transferred event for the entity and alerts the input
node that this transfer process has completed.

Figure 20.4: Station Activity

· Next add a Delay step in similar fashion as before (see Chapter 19) specifying the DelayTime as a
referenced property.
· Once the entity has been delayed, it needs to be transferred out of the current DelayObject location
(station) using a Transfer step as specified in Figure 20.5. The entity is sent to external output
node to enter back into the parent model when it leaves the delay object:

Figure 20.5: Transfer Step Specifications

322 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 11: Next, we need to use the new DelayObject in our model as seen in Figure 20.6. Switch back to the
main create Model (within the Navigation window). To facilitate multiple changes as we modify the
object we will insert the delay between two nodes so we can avoid having setting the paths each time.

· Delete the TIMEPATH connecting the TNormal to the SnkExit. Insert a new TRANSFERNODE and
connect it to the SnkExit via a 10 minute TIMEPATH. Also, you will need to specify the
SetTimeClock as the Entered add-on process trigger.
· Insert a SimpleDelay object from the [Project Library] and call it DelOperation. Specify the delay
time to be four minutes to match the TDelay time. Use connectors to link the input and output nodes.

Connecto
r

Figure 20.6: SIMO Model with Simple Delay

Step 12: Save and run the model for 24 hours observing the differences between the delay object and the
delay transfer node. Set the arrival rate to be two minutes to see the differences. Note, if your entities are
upside down in the queue, then the animated queue in the “External” view was drawn in the wrong
direction (i.e., should be right to left).

Question 1: What is the visual difference between the DELAYTRANSFERNODE and the SIMPLEDELAY with
regard to flow of entities?
________________________________________________________________________________

Step 13: Our new SIMPLEDELAY object acts more similar to the standard SERVER object now with entities
shown delaying in the station Contents queue. To compare our new object to the SERVER, replace the
DELAYTRANSFERNODE with a SERVER named SrvComparison

· Set the Processing Time and Delay Time to Random.Pert(1,2,4,1) minutes for both the
SrvComparison and the DelOperation respectively.
· Set the SrvComparison’s Initial Capacity to “Infinity” as seen in Figure 20.7.
· Connect the ParentOutput@SepCopy to the SrvComparison via a five minute TIMEPATH.
· Connect the SrvComparison to the SnkExit via a ten minute TIMEPATH.
· Also, connect the SrcParts directly to both the SrvComparison and DelOperation via CONNECTORS.

Creating New Objects 323

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Use “Specific” destination 10 min
TimePath

Interarrival:
Exponential(2,2) Connecto
r
Figure 20.7: Comparing the SimpleDelay with the Server
Step 14: Change the Interarrival Time of the SrcParts to Random.Exponential(2, 2)292To allow us
to do a comparison of the system, use a “Specific” as the Entity Destination Type for routing from the
SOURCE output and specify the Input@SepCopy.

Step 15: Save and run the model for 100 hours.

Question 2: When you execute the model for 100 hours, can you see any difference in behavior or results?
________________________________________________________________________________

Step 16: So we can run experiments with different destinations, define a Node Object Reference property
called DestinationNode whose Default Value is Input@DelOperation. Change the Entity
Desitination Type from the Output@SrcParts is based on a specific destination using the node reference
property.293

Figure 20.8: Specifying the Use of DestinationNode Property

Step 17: Remove the Entered Add-on process triggers for setting the time clock from the time path heading
to the DelOperation, the time path leaving the DelOperation and the TNormal TRANSFERNODE.

Step 18: Create a new Experiment and set the parameters to one replication of 1000 hours. Run the
experiment with the DestinationNode control set to new delay node (Input@DelOperation) and then rerun
it again with the destination being the infinite capacity server.

Question 3: What did you get for the execution time for each of the experiments? Do you think this
difference in time is important?
________________________________________________________________________________

Question 4: What other factors influence this execution time comparison?

________________________________________________________________________________

292
Note we are using the second parameter of the Exponential to specify the specific random number stream.
293
Again you will need to type in the reference since it is not shown in the drop-down list.
324 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 5: Although execution time is better for a SimpleDelay object versus a SERVER object, is
execution time the most important consideration when creating a simulation object?
________________________________________________________________________________

Part 20.2: Adding Color and States

One of the features of the SIMIO objects (e.g. SERVER) is they change color based on their status (state). With
some work we can mimic this behavior in our delay object, so it turns a different color when entities are being
delayed versus when the delay object is idle. Furthermore, we can also update their state and automatically
generate statistics on the object’s use. We begin with the state determination.

Step 1: The SERVER keeps track of its states using a LIST STATE variable named “ResourceState”. The
LIST STATE VARIABLE defines a discrete integer variable with a list of possible values from zero to N.
Select the SIMPLEDELAY from the [Navigation] panel. Within the new object, first define a string list
with the names. From the Definitions→Lists, insert a new string list named LstDelayStateNames with
“Idle” and “Delaying” as the two potential states.

Figure 20.9: Specifying the Names of the Potential States

Step 2: Insert a new LIST STATE variable named StaDelayState which uses the LstDelayStateNames list.
Since SIMIO will automatically track statistics on each state, specify the Data Source and Category
property as seen in Figure 20.10.

Figure 20.10: Setting up the List State Variable.

Step 3: Once the entity enters the delay station, the delay state will be delaying and will potentially become
idle when it leaves the simple delay object as seen in Figure 20.11.

Figure 20.11: Modifying the OnEnteredDelay to Change the Delay State

· Insert an Assign step that sets the StaDelayState to “1” immediately after the End Transfer.

Creating New Objects 325

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 · Right before the entity leaves (i.e., Transfer step), we need to update the status to idle if this is the
only entity in the delay station. Insert another Assign step that sets the StaDelayState variable to
StationDelay.Contents.NumberWaiting > 1 which will evaluate to zero if this is the last
entity in the delay station or one as seen in Figure 20.12.

Figure 20.12: Changing the State of the SIMPLEDELAY Object

Step 4: Return back to the main Model and set the arrival destination node to Input@DelOperation.

Step 5: Save and run the model for 24 hours looking at the results.

Question 6: What percentage of the time was the delay object idle and delaying?
________________________________________________________________________________

Question 7: While the simulation was running could you tell when the DelOperation was delaying?
________________________________________________________________________________

Step 6: Currently there is only one symbol defined. To add additional symbols to our simple delay, click
first on the “External View” for the SIMPLEDELAY object and select the Additional Symbols tab. Now add
an additional symbol and color the ellipse of the new one “green” (i.e., the green color will represent the
status “Delaying”, meaning there are entities in the delay). See the symbols in Figure 20.13.

Figure 20.13: Symbols for the Two States Idle and Delaying
Step 7: To use the states to define which symbol to display, change the Default Current Symbol Index in
the Properties window to the SimpleDelay.StaDelayState.

Figure 20.14: Using the Delay State to Change the Symbol

Step 8: Return to the main Model and delete the current DelOperation object from your SIMIO model
since the external view is not updated with instances already defined. Insert a new SIMPLEDELAY using the
same name, making sure to set the delay time to Random.Pert(1, 2, 4, 1) minutes. Save and run the model
interactively.

326 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 8: Does it change color when entities occupy the delay?294
________________________________________________________________________________

Question 9: What other embellishments to this simple delay object do you think would be useful?
________________________________________________________________________________

Part 20.3: Adding Defined Add-on Process Triggers

Perhaps the greatest flexibility you can add to your objects is to allow your users to add on processes that
change the behavior of the object to meet their specific needs. This addition does assume your user will know
how to create those processes. Many of the SIMIO standard objects allow the user to augment the processing
of an object via add-on process triggers (e.g., Run Initialized, Entered, Exited, Processing, Processed, etc). To
allow our new object to mimic the standard objects, these user defined add-on process triggers need to be
specified.

Step 1: Select your SIMPLEDELAY from the [Navigation] panel and navigate to the Definitions→ Properties
section. Properties associated with the add-on process triggers have to be defined so the user will be able
to specify them. Five process triggers will be added (Initialized, Entered, Exited, BeforeDelaying, and
AfterDelaying) using a “Process” property under the “Element Reference”.

Step 2: Insert a PROCESS ELEMENT REFERENCE property for the “Run Initialized” with the following
properties. Like the other objects we will have this process run at the start of a simulation. New category
names can be created by just typing in the category into the entry box.

Property Value
Name RunInitializedAddOnProcess
Description Executes when the simulation run is initialized.
Required Value False
Display Name Run Initialized
Category Name Add-on Process Triggers
Default Value Empty String
· Add the “Entered” PROCESS ELEMENT REFERENCE property with the following properties that will be
executed as soon as an object enters the DELAYOBJECT.

Property Value
Name EnteredAddOnProcess
Occurs immediately after an entity has entered this
Description
object and before the delay.
Required Value False
Display Name Entered
Category Name Add-on Process Triggers
Default Value Empty String
· Add the “Exited” PROCESS ELEMENT REFERENCE property with the following properties that will be
executed once an object leaves the DELAYOBJECT.

294
Note, the user has the ability change the Current Symbol Index in the model just like any other standard SIMIO object.
Creating New Objects 327

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Property Value
Name ExitedAddOnProcess
Display Name Exited
Occurs immediately after an entity has exited this
Description
object
Category Name Add-on Process Triggers
Required Value False
Default Value Empty String
· Add the “Before Delaying” PROCESS ELEMENT REFERENCE property that will run immediately before
the ENTITY starts the delay process similar to the SERVER’S Before Processing trigger.

Property Value
Name BeforeDelayingAddOnProcess
Display Name Before Delaying
Description Occurs immediately before an entity is to be delayed
Category Name Add-on Process Triggers
Required Value False
Default Value Empty String
· Finally, insert the “After Delaying” PROCESS ELEMENT REFERENCE property that will run once the
ENTITY has finished delaying similar to the SERVER’S After Processing trigger.

Property Value
Name AfterDelayingAddOnProcess
Display Name After Delaying
Description Occurs immediately after an entity has been delayed
Category Name Add-on Process Triggers
Required Value False
Default Value Empty String
Return to the Processes tab and insert an Execute step into the OnEnteredDelay process which allows
processes to be executed as seen in Figure 20.15. Specify the process to be the referenced property
(EnteredAddOnProcess). The Token Wait Action property should be “WaitUntilProcessCompleted”
under the Advanced Options which forces the token to wait until the add-on process has completed running
before it continues to the next step.

EnteredAddOnProces AfterDelayingAddOnProcess
s

BeforeDelayingAddOnProce
ss

Figure 20.15: Specifying the Entered and Delayed Add-on Process Trigger
Step 3: Before the Assign step (see Figure 20.15) that changes the state in the OnEnteredDelay process,
insert another Execute step that will run the process associated with the
BeforeDelayingAddOnProcess trigger with the same (default) Action property value.

Step 4: After the Delay step, insert another Execute step that will run the process associated with the
AfterDelayingAddOnProcess trigger with the same (default) Action property value.

328 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 5: Add the OnRunInitialized Process which is automatically defined for all FIXED CLASS objects by
selecting it in the “Select Process” dropdown box. Insert an Execute step into the process as before but
specify the RunInitializedAddOnProcess referenced property as seen in Figure 20.16 and Figure 20.17.

Figure 20.16: Using a Built-in Process for all Fixed Objects

Step 6: Create a new Process named OnExited and specify the TriggeringEvent property to be the
StationDelay.Exited event. Again, insert an Execute step that will execute the process for the
user defined ExitedAddOnProcess property.

Figure 20.17: Adding an OnRunInitialized and OnExited Process

Step 7: Let’s utilize our new add on process triggers, as a “user” of our object, to allow the delay time to
vary for each entity (i.e., 25% of the time it will be four minutes, 25% of the time it will be six minutes
and 50% of the time it will be eight minutes).

· Insert a new MODELENTITY real valued state variable named EStaDelayTime whose Unit Type is
Time with default units of “Minutes” which will be used to store the delay time.
· Specify the Delay Time of the DelOperation be ModelEntity.EStaDelayTime.
· Navigate back to the Model and create a new Entered Add-on Process trigger for the DelOperation as
seen in Figure 20.18.
· Add an Assign step that assigns ModelEntity.EStaDelayTime a value from a Random.
Discrete(4, 0.25, 6, 0.5, 8, 1.0) minutes.

Figure 20.18: Assigning a Delay Time for Each Entity

Step 8: Save and run the model for 100 hours designating all entities to the new object.

Creating New Objects 329

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 10: Theoretically, the time in station should be 6.5 minutes (why?). What did you get for the
“Holding” time at the delay station in DelOperation?
_______________________________________________________________________________

Question 11: How important might the add-on processes be in a simple object?
________________________________________________________________________________

Part 20.4: Embellishing with State Assignments

The SIMIO standard objects have state assignments that can perform assignments without the use of
processes. This addition is considered a “convenience” to the user, especially since assignments are easily
implemented in processes. Nonetheless, similar state assignments can be added to the new SIMPLEDELAY to
illustrate how to add this capability. Recall that an Assign step may consist of several assignments. As a
result, this addition must accommodate repeating specifications.

Step 1: State assignments are specified by a “repeating group property” to the new object so the number of
assignments can vary. Add a Repeat Group property named AssignmentsOnEntering in the
SIMPLEDELAY object with the properties specified in Figure 20.19. Again, type in the new category name
State Assignments so it will be similar to the other standard objects. The repeat group allows you to
specify multiple rows (i.e., repeating) of a set of properties (i.e., the state variable and the value
expression).

Figure 20.19: Adding a Repeating Group Property

Step 2: To create the properties that are to be repeated (i.e., grouped), first select the Assignments
OnEntering group property, and then insert a new State Standard Property named
AssignmentsOnEnteringStateVariableName since a state variable is used in an assignment with the
properties specified in Figure 20.20.

330 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 20.20: Specifying a State Variable Property

Step 3: Next, insert the expression property that will specify the new value to be assigned the state
variable. Again, select the AssignmentsOnEntering group property first, and then insert a new
Expression Standard Property named AssignmentsOnEnteringNewValue with the properties specified
in Figure 20.21. Note carefully the specification of the Unit Type, so units of the value can be specified
based on the state variable.

Figure 20.21: Specifying the Expression Value Property of the Repeating Group

Step 4: Repeat the process for another Repeat Group Property named AssignmentsBeforeExiting as seen
in Figure 20.22 which shows all the properties. Notice how the group properties have their own separate
properties section. You can copy the AssignmentsOnEntering and just change the name.

Creating New Objects 331

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 20.22: Properties Associated with the SIMPLEDELAY

Step 5: Once the Repeating Group properties have been defined, they can be used in Assign steps. Insert
two Assign steps as seen in Figure 20.23 before the “OnEntering” and “BeforeExiting” state
assignments. One Assign step should be inserted before the execution of the entered add-on process and
the other one right before the entity is transferred out of the node. Figure 20.23 shows how the reference
properties are used to specify the Before Exiting Assignments property repeating group.

Figure 20.23: Adding the Assign Process Steps for the Assignment States

Step 6: For each of the Assignments (More) properties, the state variable and new value properties have to
be associated with the reference properties as well. Bring up the Repeating Property Editor by clicking the
more button as seen in Figure 20.24.

332 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 20.24: Setting up the State Variables and New Values to the Referenced Properties
Step 7: For both the State Variable Name and New Value properties, set the reference property to the top
property for each property. Figure 20.25 shows selecting the State Variable Name referenced property.

Figure 20.25: Select the Top Reference Property for the State Variable Name
Step 8: Let’s utilize our new processes and assignments by specifying the delay time of each of the entities
using the OnEntering State Assignments.

· First remove the Entered Add-on process trigger by resetting it (reset or Null) so the delay time will
not be specified by the add-on process.
· Repeat the same assignment as was done in Figure 20.26 using the AssignmentsOnEntering State
Assignments. Assign the ModelEntity.EStaDelayTime a value from a Random.Discrete(4,
0.25,6,0.5,8,1.0)minutes.295

Figure 20.26: Using the State Assignments Property

Step 9: Save and run the model for 100 hours designating all entities to the new object.

Question 12: What did you get for the Holding time in the DelOperation?
________________________________________________________________________________

Question 13: How important are the state assignments in a simple object?
________________________________________________________________________________

Question 14: Is the state assignment needed when there are add-on triggers?
________________________________________________________________________________

295
If the Units do not show, then the New Value property was not setup correctly (see Figure 20.21).
Creating New Objects 333

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Part 20.5: Capacity and Its Complications

Adding capacity to our delay object adds new complications because entities cannot enter a delay when the
delay object is at full. Something needs to be done with the arriving entities

Step 1: Rename our SIMPLEDELAY object296 to CAPACITYDELAY and change the “Resource Object” property
in the “Advanced Options” to “True”. We will use the capacity specifications that are a consequence of
making this a resource object (note the “additions” to the “Process Logic” category).

Step 2: Set the Initial Capacity of the StationDelay station to a reference to InitialCapacity so that the
capacity of the station is identical to the CAPACITYDELAY object.

Figure 20.27: Setting Station Capacity

Step 3: The OnEnteredDelay process can now be modified as shown in Figure 20.28. To seize capacity
from the object, use PARENTOBJECT as the Object Type within the Seize step. The same object type will
be used at the Release step. We will use only one unit of capacity from this object for any delayed
entity. We will choose to put the Seize step before the change in state assignment and the Release
step directly after the Delay since we know the object is no longer busy

Seize Step Release Step

Figure 20.28: New CapacityDelay Process

Step 4: Use this in the Model by replacing the SimpleDelay object with the CapacityDelay object. Run
the model with the PATH to the MyCapacityDelay having “no passing” Run it again to “allow passing”
to that path297.

Question 15: What happens to the entities that can’t enter the object’s station?
________________________________________________________________________________

Step 5: When you run the model, say for 24 hours, ignoring the build-up in the path, examine the
“Results”. Look at the output of the StationDelay and notice there are now queuing statistics on the
EntryQueue of the station.

Step 6: The EntryQueue is a SIMIO defined queue that exists for each station, in case the station is full
when the entity arrives. Thus the station element has two queues associated with each station: Contents

296
You could start at the beginning and create a new Fixed Class object.
297
The connections need to be paths to have the “no passing” option. Other types of paths do not have this feature.
334 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 and EntryQueue. (If you specify that the input buffer is ten for a SERVER and the eleventh entity arrives to
the server, the entity will be placed into the EntryQueue and not be allowed to enter the server - it will stay
at the input node of the server).

Step 7: Now let’s modify our object so that the EntryQueue is visible as shown below in Figure 20.29.
Then replace the existing CapacityDelay with this new CapacityDelay and execute the model. Use a
capacity of 1 and a delay time of 2 minutes. Observe the behavior of the entities entering the station.

EntryQueu
e

Figure 20.29: Add Entry Queue

Question 16: Is there anything about the animation or the statistics that trouble you?
_____________________________________________________________________________

Step 8: Although the statistics are ok, the animation shows an entity waiting on the path that is duplicated
within the visible entry queue. This duplication occurs because the delay station is full and there are
entities waiting.

Step 9: You will recall that the standard SERVER object uses a station for waiting – namely the InputBuffer.
So let’s use a second station in our object to account for waiting to enter the delay.

· Insert a new STATION named StationWaiting. Let’s make its Initial Capacity Infinity so there is
no limit on waiting especially since we do not want to deal with another set of capacity concerns.
· In the External view change the Input External Node to reference the StationWaiting instead of the
StationDelay to route entities to this station first.
· Change the queue representing the EntryQueue to the StationWaiting.Contents queue.

Step 10: Finally, we need to insert an OnEnteredWaiting process to handle entering the waiting station.
Move the Assign, Execute, and Seize steps from the OnEnteredDelay to the OnEnteredWaiting.
Doing this will separate the time spent waiting for capacity from the time in the delay – see Figure 20.30.

OnEnteredDelayed
OnEnteredWaiting

Figure 20.30: OnEntered Processes

Step 11: Configure the DelOperation in the Model to have a capacity of “1” and a delay time of “2”
minutes. Run the model interactively to observe the behavior of the new object.

Creating New Objects 335

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 17: Does the observed queuing seem appropriate?
_____________________________________________________________________________

Step 12: Now run the Simulation for 100 hours.

Question 18: What is the average time (in minutes) spent waiting to enter the station?
_____________________________________________________________________________

Question 19: What is the average holding time spent in the station (in minutes)?
_____________________________________________________________________________

Part 20.6: Adding Multiple Resources

The capacity of the new object would be greatly enhanced if multiple resources could be required before the
entity is delayed. We would to like require resources before the delay occurs and then release them right after
the delay, as is done with Secondary Resources in the SERVER. Again a Repeat Group property similar to the
assignments has to be utilized since the Seize and Release steps within the secondary resources can
request/release more than one type of capacity.

Step 1: Recall, the assignments had just two properties (i.e., the state variable and the new value)
associated with the Repeat Group. That is not the case for the Seize and Release steps which have
many properties. The easiest way is to sub-class the SERVER object and then copy and paste the secondary
resources (e.g., SecondaryResourcesSeizesBeforeProcessing and
SecondaryResourcesReleasesAfterProcessing) into the SIMPLEDELAY. Therefore, sub-class the SERVER by
right clicking to create a MYSERVER and then expand the “Inherited Properties.”

Step 2: Select the MYSERVER from the [Navigation] window. Under the Definitions→Properties section,
expand the Properties(Inherited). Copy the SecondaryResourceSeizesBeforeProcessing and then paste
this Repeat Group property into the CAPACITYDELAY properties.

Step 3: Repeat the process for the SecondaryResourceReleasesAfterProcessing Repeat Group property
(see Figure 20.31).

Figure 20.31: Repeat Group Properties for Allowing Secondary Resources

Step 4: From the OnEnteredProcessing process (see Figure 20.32) inside the MYSERVER object, copy the
Seize and Release steps before and after process.

336 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Copy Release Step
(After Processing)

Copy Seize Step

(Before Processing)

Figure 20.32: Copy Seize and Release Steps from the MyServer’s OnEnteredProcessing Process
Step 5: Now in the OnEnteredWaiting process of the SIMPLEDELAY object, add the Seize secondary
resources before the Seize step which seizes the parent object as in Figure 20.33.

Add Seize (before Transfer)

Figure 20.33: Adding the Seize into the OnEnteredWaiting

Step 6: Next in the OnEnteredDelay process, add the Release right after the Delay step as seen in
Figure 20.34.
Add Release (after Delay)

Figure 20.34: Adding the Release to the OnEnteredDelay

Step 7: Return to the Model and insert a RESOURCE named ResService with a capacity of two, of which
one unit is required before the delay can occur. The DelOperation also has a capacity of one (it only
makes sense for the delay capacity to be one or two, since the secondary resource has capacity of two).

Step 8: In the “Secondary Resources” of the DelOperation, seize the ResService in the Other Resource
Seizes→Before Processing and release it in the Other Resource Releases→After Processing.

Step 9: Save and run the model interactively observing the animation.

Question 20: Does the animation behave as you expect?

________________________________________________________________________________

Creating New Objects 337

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 21: For a simulation of 100 hours, what is the holding time in the StationDelay and in the
StationWaiting
________________________________________________________________________________

Question 22: What is the scheduled utilization of DelOperation and the ResService? Why is the utilization
one half of the other?
________________________________________________________________________________

Part 20.7: Using Storages to Distinguish Waiting versus Delaying

The processing queue for the delay station now includes waiting for secondary resources and the actual delay,
so it’s holding time increases. We mimicked the SERVER object and utilized two stations (i.e., StationWaiting
and StationDelay) to decouple the waiting and processing times. However we have two stations and two sets
of entering processes. Furthermore we have two sets of Seizes and Releases for secondary resources and
the actual delay.

To simply our object, let’s abandon the capacity limit on the delay. By doing so, we limit the resources to
those specified as “secondary resources” or in process add-on triggers. In essence these will be our only
resource specifications and in turn we have a simpler object. Further, rather than two stations (locations)
within the object, it would be better to have one station, so we avoid the complications of transferring between
stations (these transfers produce events in the current events calendar, meaning they may not be executed
immediately). Instead we will utilize STORAGE elements which also automatically calculate statistics and
which we have used earlier.

Step 1: Select the CAPACITYDELAY object from the [Navigation] panel. Right-click to bring up its
“Properties” and under the “Advance Options” make the Resource Object False. This means our object
will no longer have capacity limitations.

Step 2: Since the StationWait will be removed, reset the OnEnteredDelay process as seen in Figure
20.35.

· Move the “OnEnteringAssignment” Assign and the “EnteredAddOnProcess” Execute from the
OnEnteredWait before the End Transfer.
· Move the “SecondaryResources” Seize and the “BeforeDelayingAddOnProcess” Execute from
the OnEnteredWait right after the End Transfer.
· Finally, remove the Release step that releases the parent object in the OnEnteredDelay.

Figure 20.35: Resenting the OnEnteredDelay

Step 3: In the Definitions→ Elements section, delete the StationWaiting station. Next, insert two
STORAGE elements named StorageWait and StorageDelay as in Figure 20.36. Finally, set the
InitialCapacity of the StationDelay back to Infinity so the station has no capacity limit.

Figure 20.36: Adding Two STORAGE Elements to Keep Track of Waiting and Delaying

338 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 4: To calculate how long entities waited for secondary resources, we need to insert the entities into the
StorageWait before we Seize and then Remove them after the BeforeDelaying procces-trigger. We
need to repeat the process for keeping track of the actual delay time by inserting the entities into the
StorageDelay right before the Delay and then removing them right after the Delay as seen in Figure
20.37
Insert/Remove Insert/Remove
StorageWaitQ StorageDelayQ

Figure 20.37: Utilizing the StorageWait and StorageDealy to Track Waiting Times
Step 5: Finally modify the CAPACITYDELAY object “External” view. Insert a new animated queue that
references StorageWait.Queue while the processing queue should reference the StorageDelay.Queue.
Change the Input “External Node” to send entities entering (i.e, its Station Name is StationDelay).

StorageDelay.Queue
StorageWait.Queue

Figure 20.38: Adding the Delay and Wait Queues to the External View
Step 6: Save and run the Main model for 100 hours. Be sure your model seizes one unit of ResService
before processing releases one unit of ResService after processing. The delay time is 3 minutes. You will
need to delete the current CapacityDelay and reinsert it for the new external view to be used.

Question 23: When you switched back to the Main model, are the two new queues visible?
________________________________________________________________________________

Question 24: On average, how long did parts wait for the MyCapacityDelay?
________________________________________________________________________________

Question 25: How long were parts actually delayed in the MyCapacityDelay?
________________________________________________________________________________

Question 26: Comment on the simplicity and transparency of this object.

________________________________________________________________________________

Step 7: At this point, we have replicated a lot of the SERVER object, which is a complex object (i.e., a heavy
weight object), versus our CAPACITYDELAY (i.e., light weight) object. Therefore, we will check our
created object with that of the SERVER again.

· For the SERVER SrvComparison, make sure the Initial Capacity to “Infinity” and set the Processing
Time property to ModelEntity.EStaDelayTime.
· For the DelOperation, set the Delay Time property to ModelEntity.EStaDelayTime.

Creating New Objects 339

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 8: In the SrcParts_CreatedEntity add-on process, add the Assign step to set the entities delay time
to Random.Pert(1,2,4,1)298, as in Figure 20.39.

Figure 20.39: Delay Time Assignment

Step 9: Next for the SrcParts, set the Interarrival Time to Random.Exponential(7,2).

Step 10: To compare the two systems, we need to add the use of secondary resources to both objects. We
will require one unit of ResService at each.

· In the “Secondary Resources” of the CapDelayComparison, seize one unit of the ResService in the
Other Resource Seizes→Before Processing and release it in the Other Resource Releases→After
Processing.
· In the “Secondary Resources” of the SrvComparison, seize one unit of the ResService in the Other
Resource Seizes→Before Processing and release it in the Other Resource Releases→After Processing.

Step 11: Set the DestinationNode to Input@SepCopy and run the simulation for 1000 hours.

Question 27: What did you get for the TimeWaiting in the StorageWait of the CapDelayComparison and
the HoldingTime for the InputBuffer of the SrvComparison?
________________________________________________________________________________

Question 28: Why are these identical?

________________________________________________________________________________

Question 29: What corresponds to the TimeWaiting in the StorageDelay of the CapDelayComparison and
relative to the SrvComparison?
________________________________________________________________________________

Step 12: Employ a SIMIO experiment where the DestinationNode is the only control. First run the
experiment for 1000 minutes with the DestinationNode being Input@DelOperation. Then run the
experiment with DestinationNode being Input@SrvComparison.

Question 30: What did you get for the execution time for each of the experiments? Do you think the
difference is important?
________________________________________________________________________________

Step 13: Now in the Experiment, set up two scenarios that specify Input@DelOperation and
Input@SrvComparison respectively with 100 replications. Change the run length back to 100 hours.

Step 14: Save and run the experiment.

298
Note we are using the fourth parameter which specifies the random stream so we induce as much correlation into comparing
scenarios.
340 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 31: What is the average and half width of total time in system as well as the time waiting for the
two scenarios?
________________________________________________________________________________

Part 20.8: Some Observations on the Design of Objects

In this chapter we carefully and deliberately created an object from one of the SIMIO base classes. Since the
goal was to create a “queuing-like” object, our SIMIO reference was the SERVER object. But we started where
the SIMIO developers may have started, namely with a base class. Our initial object was a simple delay based
on a station, but we were able to show how the object could be enhanced with assignments and add-on
processes. In the midst of our design concerns, we introduced capacity into the simple object through the
station and that brought several queuing complications, which the SIMIO developers also had to resolve for
the SERVER. Although we considered several options, we eventually abandoned the capacity of the station in
favor of simply using “secondary resources” as the primary resources. And we obtained statistics on waiting
and in delay from storages.

It is useful to contrast the simplicity and power of the created object with the SERVER. Without doubt the
SERVER provides a wide range of powerful features. For example the SERVER has a variety of queuing options,
it has input and output buffer capacity, and the processing can handle tasks and task sequences. There is
reliability logic and financials associated with the SERVER and the new object doesn’t have buffer locations for
either input or output.

Nevertheless, the new object has the benefits of simplicity and transparency. The SERVER has three stations
and numerous processes, including three entering process. The new object has only one station and one
entering process. Further, looking at the SERVER processes, notice the proliferation of steps, add-on processes,
and assignments. This proliferation creates a challenge to understand just when something is done in the
SERVER and further, it now becomes a challenge as to when each of these features are invoked. The new
object features are much clearer The fact there is only one entering process, it is easy to see when add-on
processes, assignments, and resources seize/release occur. Also in the SERVER, the processes must account for
when the stations have zero capacity. In the new object, the only resource constraint are the secondary
resources.

While stations provide the possibility of a capacity constrained location, the capacity must be seized and
released while the location is obtained by transfer. Leaving a station fires a transfer event that must be
resolved from the current events. In fact, these current events emanate from the set of processes in the object.
If you examine the SERVER carefully, you will notice the many processes which cause any number of current
events and we have seen in prior examples how these “zero-time” events can dramatically affect the
simulation execution.

So our advice relative to creating you own objects is to focus on simplicity and transparency so using your
object does not become an error-prone challenge. The easier it is to use an object and the more transparent its
behavior, the more success you will have in creating object-oriented simulations

Part 20.9: Commentary

· Creating your own specialized objects can really enhance your models, which is one of the important
advantages of using SIMIO. However, if you are only going to use the object once and in one
particular model, it may not make sense to spend the time creating a specialized object that could be
handled in processes. The advantage is the reuse with-in the same model or different models.
· Instead of the need for a capacity on the station (location), considered the capacity restriction of the
object as being a separate concern from the resources needed. It greatly simplified the object,
especially in contrast to the SIMIO server.
Creating New Objects 341

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 · But as we developed the object we realized that there is a strong interest to adding “features” and the
appropriate stopping point is not at all clear.

342 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Chapter 21
Continuous Variables, Reneging, Interrupt, Debugging: A
Gas Station
Continuous variables are variables whose value can change continuously during the simulation (as opposed to
discrete variables). Continuous variables in SIMIO can be “Level” or “Level with Acceleration” state
variables – see “States” under the Definitions”tab in the “Continuous” section. When a LEVEL state variable is
created, you may also specify the rate of change, called the Initial Rate Value property as well as its initial
state called the Initial State Value property. LEVEL WITH ACCELERATION variables in addition to specifying
the rate of change and the initial state, you may also specify acceleration rate, called Initial Acceleration Value
property as well as the length of the acceleration via the Initial Acceleration Duration Value. Unlike discrete
state variables, continuous state variables can change continuously based on the rate and acceleration.

Part 21.1: Simple Tank299 Manual Process

We are going to manage a simple gasoline tank by changing its rate value. We will assume the tank volume is
between 0 and 100 gallons.

Step 1: Create a new model. Add a fixed RESOURCE object to the model named ResTank which will
represent the tank.

Step 2: Create a new symbol named TankSymbol to associate with the tank, (from the Project
Home→New Symbol→Create Symbol) such as the one in Figure 21.1. The symbol which represents the
interior is a square with a height of ten meters (whose height determines the content/level of the tank).
Add a Gold Comb texture to give it the appearance of gasoline for both the top and the sides.

Figure 21.1: 2D and 3D Representation of the Tank Symbol

Step 3: Next, use the TankSymbol as the symbol for the ResTank RESOURCE.300 Once this is done, draw a
static “case” around it so you can see the level of the tank in 3D as shown in Figure 21.2. The case is
made from a thick “Polyline”, leaving a small opening in the front, so as to see the level of the tank. The
polyline also needs a height of ten meters.

299 This “simple tank” example is adapted from the SIMIO “SimBits” which are another way help you learn about SIMIO.
300
Select the RESOURCE object and then choose the symbols from the “Symbols” tab.
Continuous Variables, Reneging, Interrupt, Debugging: A Gas Station 343

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 21.2: Tank Symbol
Step 4: Now add a LEVEL continuous state variable named GStaLevelConTank from the
Definitions→States section. Let’s set the Initial State Value to 0 and the Initial Rate Value to 0. A
negative rate value means the tank is being drained while a positive value means it is being filled.

Step 5: When we need to modify a parameter for a model, it is convenient to specify a model property
rather than specifying a constant. Doing so also helps in eliminating mistakes by using the same symbolic
name for a particular value. Therefore define a minimum and maximum level for the tank as a standard
real property named TankMin and TankMax.301 For these two properties, the Unit Type should be
“Volume” and Default Units equal to “Gallons.” Add two “Expression” standard properties named
RatePositive and RateNegative to represent the rate at which the tank will fill up and the rate at which the
rate will be emptied. For these two properties, the Unit Type should be “VolumeFlowRate” and Default
Units equal to “Gallons per Hour.” Place each property into a new Tank Category.

Step 6: Next, right click on the Model in the [Navigation] panel to access the model properties and set the
values of the minimum level property (TankMin) to 0, the maximum level property (TankMax) to 100,
the fill rate (RatePositive) to 100, and the emptying rate (RateNegative) to -200.

Step 7: Now add two MONITOR Elements named MonTankEmpty and MonTankFull. These MONITORS
will detect the crossing of the GStaLevelConTank at the TankMin and TankMax levels respectively.

Step 8: For the MonTankEmpty MONITOR, set the properties as shown in Figure 21.3.

Figure 21.3: TankEmpty Monitor Specification

Step 9: And for the MonTankFull MONITOR, see Figure 21.4 for the property settings.

301
This is done under the Definitions tab in the Properties section..
344 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 21.4: MonTankFull Monitor Specification
Step 10: Create new processes for the two On Change Detected Process add-on process triggers which are
defined in Figure 21.5 and Figure 21.6.302
Step 11: The MonTankEmpty_OnChangeDetectedProcess process is called when the tank empties (i.e., the
GStaLevelConTank reaches zero). We will need to stop emptying the tank, have a small delay before
beginning to fill, and then start filling as shown in Figure 21.5.

Figure 21.5: When the Tank Empties

· The first Assign step sets both the GStaLevelConTank.Rate to “0” which stops the LEVEL state
variable from changing and sets the ResTank.Size.Height to “0”.
· Use the Delay step to force a 0.2 hour delay before the tank begins to fill again.
· The final Assign step sets the GStaLevelConTank.Rate to its maximum value of RatePositive
which will allow the tank to start to fill since it is a positive value.

Figure 21.6: Assigning the Rate and the Height

Step 12: The MonTankFull_OnEventProcess is called when the tank fills to the maximum level. Once it
reaches this level, stop filling, force a small delay before allowing it to empty, and then allow the
emptying to start as seen in Figure 21.7.

Figure 21.7: When the Tank Fills

302
You can either create the process here that will respond to the MONITOR event or you can create a generic process with the Triggering
Event property set to the MONITOR event (e.g., MonTankFull.Event).
Continuous Variables, Reneging, Interrupt, Debugging: A Gas Station 345

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 · The Assign step sets the GStaLevelConTank.Rate to 0. A 0.1 hour delay is assumed before the
tank can start emptying and then the last Assign step will set the rate to a negative value to allow
emptying (i.e., set the GStaLevelConTank.Rate to the RateNegative).

Step 13: Finally, we need to take care of the “height” of the tank animation. We will do that in the
OnRunInitialized process for the model which will run when the simulation starts. From the “Processes”
tab, select the OnRunInitialized process from the “Select Process” drop-down.

· This process should Execute303 the MonTankEmpty_OnEventProcess once to initialize the tank for
filling. Then, it will Delay for one minute and then Assign the ResTank.Size.Height to
GStaLevelConTank /(TankMax/10) as in Figure 21.8.

Figure 21.8: Changing the Height of the Tank

· Notice, the system loops continuously every minute updating the tank height for animation purposes.
Do this by grabbing the “End” point and dropping it on top of the Delay step. Therefore the TOKEN
will continually loop through these steps until the simulation stops.

Step 14: For the tank level to show in the animation, the Current Symbol Index in the “Animation” section
of the ResTank resource must be set to zero, so the OnRunInitialized process is the only specification to
change the height. Select the ResTank RESOURCE and add an additional symbol and then set the Current
Symbol Index property to zero.

Step 15: To complete our animation, add six STATUS LABELS from the Animation tab as seen in Figure 21.9
where the expressions are GStaLevelConTank.Rate*264.1720524, GStaLevelConTank
*264.1720524, and ResTank.Size.Height respectively304. Also, for the word Status labels,
change the background to white.

Status Labels
Status Labels

Figure 21.9: Adding Status Labels

Step 16: Next, add a Status Plot of GStaLevelConTank vs. Time (hours) – see Figure 21.31. Change the
Title, x-axis label, y-axis label, text scale and the time range to plot (i.e., 2 hours) from the Appearance
tab.

303 The Execute step will run any process that has been defined.
304
The default unit for volume and flow rate is cubic meters and cubic meters per hour which we are converting back to gallons and
gallons per hour.
346 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 21.10: Status Plot
Step 17: Also, a Circular Gauge (from the “Animation” panel) may be placed on top of the tank as seen in
Figure 21.11.305 It is a way to visualize the tank level in 2D. After adding the circular gauge with the
correct style, under the Scale section’ set the Minimum property to zero, the Maximum property to 100 and
the Ticks to 11. The Expression should be the level state GStaLevelConTank .

Figure 21.11: Adding a Circular Gauge the Current Tank Level

Step 18: Save and run the model observing the status labels and plots. Also, switch into 3D to see the tank
content rise and fall.

Question 1: Do the animations and plots appear appropriate?

________________________________________________________________________________

Question 2: Why does the tank empty faster than filling up?
________________________________________________________________________________

Part 21.2: Simple Tank Revisited using the Flow Library

SIMIO provides a [Flow Library] to assist in handling pipes (i.e., flows) and tanks. The library provides a
FLOW SOURCE for creating flow entities, FLOW SINK, TANK, FLOWNODE (i.e., regulator), and FLOW CONNECTOR
used to connect FLOWNODES. We will model the simple tank again using the [Flow Library] as seen Figure
21.12

Figure 21.12: Simple Tank using the SIMIO Flow Library

305
Move to the 3D view and use the shift key to move the gauge up in space to place it directly on top.
Continuous Variables, Reneging, Interrupt, Debugging: A Gas Station 347

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 1: Insert a new fixed model named TankFlow.306 Copy the four properties (i.e., TankMax,
TankMin, PositiveRate, and NegativeRate) from the previous model and set their values as seen in
Figure 21.13.307

Figure 21.13: Setting up the Prosperities

Step 2: Insert a new FLOWSOURCE named FlowSrcFill which will be used to fill the tank.308 Select the
output FLOWNODE Output@FlowSrcFill and change the Initial Maximum Flow to RatePositive (see
Figure 21.14).

Figure 21.14: Setting the Correct Unit of the Output Flow

Step 3: Insert a new MODELENTITY named EntGas. Change the texture of the entity to be the Gold Comb
texture as before to cause the tank to fill up with gas.

Step 4: Next, add a TANK named TankGas which will hold the gasoline flowing from the FLOWSOURCE. Set
the Initial Volume Capacity and High Mark to TankMax as well as the Low Mark to TankMin. Tanks
have the ability to specify the capacity as well as have five different level marks (i.e., low-low, low,
middle, high, and high-high marks) which will trigger processes when the tank reaches these levels. You
should shrink the size of the tank animation to be smaller so you can see the tank filling and emptying
easier.

306 You can also just delete the two MONITOR elements, the three processes, the circular gauge, the ResTank and polyline from the

current model rather than copying the objects over to the new model.
307
Note, the rate negative is positive this time because of the way the flow library works.
308
The regulator inside the Output@FlowSrcFill that embodies the continuous level state variables and thus the rates as was done
before.
348 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 21.15: Setting up the 100 Gallon Tank
Step 5: Select the Output@TankGas FLOWNODE and set the Initial Maximum Flow Rate to 200 gallons per
hour which represents the emptying rate.309 Also, set the Regulator Initially Enabled to “False” to indicate
the regulator is closed and therefore not flowing (i.e., emptying) as seen in Figure 21.16.

Figure 21.16: Specifying the Tank Emptying Rate

Step 6: Add a FLOWSINK named FlowSnk to destroy the flow of gas emptying the tank.

Step 7: Connect the FlowSrcFill, TankGas, and FlowSnk using FLOWCONNECTORS.

Step 8: Copy the Tank Rate and Tank Level status labels as well as the status plot from the previous
model. Change the expressions to TankGas.FlowContainer.Contents.Volume.Rate *
264.1720524 and TankGas.FlowContainer.Contents.Volume * 264.1720524 for the
rate and tank level respectively. Let the plot expression be the same as the tank level expression.

Step 9: Save and run the model. You may need to modify the speed factor to 100 to see the tank level
rising.

Question 3: What happens when the simulation runs with regard to the tank and the two
FLOWCONNECTORS?
________________________________________________________________________________

Question 4: What did you notice about the tank rate when the tank reached the full capacity?
________________________________________________________________________________

309
Notice we do not specify it to be negative as was done in the manual process of the previous section.
Continuous Variables, Reneging, Interrupt, Debugging: A Gas Station 349

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 5: Does the tank empty once it fills?
_______________________________________________________________________________

Step 10: Once the tank fills up, the flow from the FlowSrcFill stops flowing but the tank does not empty as
the Output@TankGas regulator is not enabled. When the tank fills up we need to delay for 0.1 hours and
then open up this regulator while shutting down the regulator at Output@FlowSrcFill. Under the Tank
Level Rising category, insert the Tank Full Add-on Process Trigger as seen in Figure 21.17.310 The
FLOWNODES contain a REGULATOR element which can be opened and closed to control the flow.

Figure 21.17: Shutdown the Flow Source and then Open up the Tank Output
Step 11: Once the tank empties, we need shut down the outflow regulator of the tank, delay for 0.2 hours
and then open up the regulator at Output@FlowSrcFill to start filling. Under the Tank Level Falling
category, insert the Tank Empty Add-on Process Trigger as seen in Figure 21.17

Figure 21.18: Shutdown the Outflow from Tank and Start the Source Flow
Step 12: Save and run the model. You may need to modify the speed factor to 100 to see the tank level
rising.

Question 6: What happens when the simulation runs with regard to the tank and the two
FLOWCONNECTORS?
________________________________________________________________________________

Part 21.3: The Gas Station

A gas station has four gas pumps, which dispense the same grade of gas. All the gas pumps pump their gas
from a single tank in the ground. Cars arrive and obtain gas from any of the four pumps. If the ground tank is
depleted to 100 gallons, the pumps are closed, although any car currently being filled finishes. Any cars in
line will stay in line, but newly arriving cars will balk (i.e., leave). A tanker truck comes by periodically to
refill the ground tank. As soon as the ground tank is full (i.e., 10,000 gallons), the pumps are open again to
deliver gas.

310
You will have to type in these state variables as they will not show up in the dropdown list of possible state variables.
350 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 This problem can be approached in several ways. Again, a continuous variable (i.e., tank) will be used to
model the ground tank and we will let the cars and the tanker truck change the rates of inflow and outflow to
the tank. The cars and the tanker truck will be our active entities. The amount of gas any car needs is assumed
to be Triangular(4,9.5,25) gallons. The rate that any car pumps gas is
Triangular(125,150,200) gallons per hour. The tanker will fill the tank at 20,000 gallons per hour.

Step 1: We will modify the model of the simple tank from the previous section. Add two SOURCES, one
SERVER, three SINKS and two MODELENTITIES named EntCars and EntRefillTrucks, as illustrated in
Figure 21.19.

· EntCars arrive Exponentially every 1.5 minutes beginning at time 0.1 while the EntRefillTrucks
arrives Uniformly between 6.75 and 8.25 hours apart.
· Use CONNECTORS to link all of the objects together associated with the cars. Make the link between
SrcTanker and SnkTankerLeaves a PATH.311
· Make sure each of the sources creates the correct entity type. Also change the picture of the
EntRefilTrucks to be a tanker312. The EntCars should be given six different car pictures randomly
chosen (i.e., set the Random Symbol property to “True” under the Animation section).

Figure 21.19: Model Structure

Step 2: Define a new DISCRETE INTEGER state variable for the main Model named GStaPumpOn to
model whether the pumps are on or off. A value of one means “on” and zero means “off”. Set the Initial
State Value property to zero.

Step 3: Set the Initial Contents property of the TankGas to be TankMin.

311
The one path will allow the tanker to be seen as it fills up the tank. Also, we have two sinks for the cars to automatically collect
statistics on cars that balk and those do not.
312
You might want to make the tanker a little smaller.
Continuous Variables, Reneging, Interrupt, Debugging: A Gas Station 351

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 21.20: Setting the Initial Amount of Gas in the Tank
Step 4: For the Model, change the TankMax and TankMin properties to be 10,000 gallons and 100 gallons
respectively. By setting the low mark value to 100, we will not allow our tank to empty completely since
cars may still be getting gas when the tanker arrives. Change the PostiveRate and the RateNegative
model properties to 20,000 and Random.Triangular(100,125,175) gallons per hour to
represent the rate a car will empty the tank.

Figure 21.21: Setting the Rates and Initial Values of the Tank
Step 5: Now for the cars (i.e., MODELENTITY) let’s define two state variables to represent the amount of gas
needed and the rate that the person can pump gas. Select the MODELENTITY in the [Navigation] panel.
Insert the following two states which will be characteristics of the individual cars as they arrive.

· EStaGasNeeded is a DISCRETE REAL STATE variable with a Unit Type of “Volume” and an Initial
State Value of “0” gallons.
· EStaCarPumpingRate is a DISCRETE REAL STATE variable with a Unit Type of “VolumeFlowRate”
an Initial State Value of “0” gallons per hour.

Step 6: Since the cars arriving will set the outflow rate, we need to set Initial Maximum Flow Rate of the
Output@TankGas to “0” Gallons per hour.

Step 7: When cars are created by the source, define a process in the Created Entity add-on process trigger
as in Figure 21.22 which assigns the car characteristics.313

Figure 21.22: Cars Leaving the Source

313
Throughout this example we will use the Assign process step rather than using the State Assignments in the modeling objects.
Ultimately the Assign steps offer more generality since the State Assignments are only available on entry to and exit from the object.
352 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Where the Assign step sets the following two state variables making sure the rate is negative.

ModelEntity.EStaGasNeeded = Random.Triangular(4,9.5,25)
ModelEntity.EStaCarPumpingRate = RateNegative

Figure 21.23: Specifying Gas Needed and the Pumping Rate of Each Car
Step 8: Set the Processing Time property (in Hours) in the SrvPumpGas server to be the following.

ModelEntity.EStaGasNeeded/ModelEntity.EStaCarPumpingRate

This expression takes into account the amount gas needed divided by the rate the customer can pump it.
Make sure to change the Units to “Hours” and do not use the default of “Minutes”. Be sure that the Initial
Capacity property of SrvPumpGas server is set to “4” to indicate room for four cars.314

Step 9: If the pumps are on, cars will be routed to the pumps and, if they are off, they will be sent directly
to the SnkCarsNoGas. The selection link weights on the paths leaving the source are defined as follows
to ensure that one path will have a weight of one and the other one a zero.

· For the connection to SnkCarsNoGas, set the Selection Weight property to 1 - GStaPumpOn.
· For the connection to SrvPumpGas, the Selection Weight should be GStaPumpOn.

Step 10: All discrete event simulation languages use events and the user can define their own events in
SIMIO. Under the “Definitions” tab add a new EVENT named EvntStartPumps, which will allow entities
to wait until the pumps are started up after they have been turned off to start pumping again.

Step 11: Consider what happens when the tanker truck arrives, as shown in Figure 21.24. The tanker truck
will turn off the pumps and start filling the tank at 20,000 gallons per hour. The tanker truck will continue
to fill until the tank is full (i.e., the monitored MonTankFull reaches its threshold). Use the Entered add-
on process trigger for the input node of the SnkTankLeaves (i.e., when the truck entity enters the sink).315

Figure 21.24: Tanker Truck Leaves

314
You may want to arrange the processing station animated queue to be oriented with four points around the server like a gas pump.
315
It is necessary to invoke this process after the truck leaves the output at the source so that a time in system can be computed for the
truck. If this process is invoked before the truck leaves the output, then its time in system is not updated.
Continuous Variables, Reneging, Interrupt, Debugging: A Gas Station 353

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 · In the Assign step, set the GStaPumpOn state variable to zero, turn off the flow REGULATOR of the
TANK and turn on the REGULATOR associated with the FLOWSOURCE as seen in Figure 21.25.

Figure 21.25: Setting the Pump to Off and Filling the Tank
· The Wait step delays the truck until the tank is filled. The full tank will be signaled by an event
called EvntStartPumps.

Figure 21.26: Waiting for EvntStartPumps to Continue

· Once the tank has filled up we need to turn the pumps on and shut down the FLOWSOURCE REGULATOR
as well enable the tank regulator to start again.

Figure 21.27: Turning On Pumps and Shutdown Filling the Tanks

Step 12: When the pumps are off new cars that arrive will leave the system. However, cars that are currently
in line and/or filling up will remain. Those that are currently filling up will finish but no other car should
be allowed to start filling. Therefore, within the SrvPumpGas server, consider first when the processing
of a car filling their individual tank begins (see Figure 21.28). Insert the Before_Processing Add-On
Process Trigger with the following logic.

Figure 21.28: Processing Begins at the Pump

· When the Before_Processing add-on process trigger starts: Decide if GStaPumpOn == 1
354 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 · If the pumps are off (i.e., “False” branch), then “Wait” for the event EvntStartPumps.
· Otherwise, if the pumps are on, enable the TankGas FLOWREGULATOR (i.e., Output@TankGas.
FlowRegulator.Enabled = True and change the tank empting rate via an Assign step by
adding the outflow due to the car being filled to the current rate:
[email protected] =
[email protected] +
ModelEntity.EStaCarPumpingRate

Step 13: When the car finishes (i.e., processing) at the pump (i.e., After_Processing add-on process trigger
for the SrvPumpGas), refer to Figure 21.29, we need to remove the outflow due to the car being filled
(i.e., reduce rate). Use the Math.Max to ensure that you never go below zero.

[email protected] = Math.Max(0,
[email protected] -
ModelEntity.EStaCarPumpingRate)

Figure 21.29: Car Finishes at Pump Decrease Rate

Step 14: Next, take care of the process when the Tank reaches the low mark of 100 gallons. For the
TankGas, insert the TankGas_TankLevelFallingBelowLowMark add-on process trigger and an Assign
step that will set the GStaPumpOn to “0”.

Figure 21.30: Turn off the Pumps so Cars will Balk

Step 15: For the TankGas_TankEmpty (see Figure 21.32) we need to disable the tank’s out flow (i.e., set
[email protected] to “False) and then set the rate to zero (i.e.,
[email protected]=0.0).

Figure 21.31: Tank is Empty

Continuous Variables, Reneging, Interrupt, Debugging: A Gas Station 355

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 16: And for the TankGas_TankFull event (see Figure 21.32), it should turn the pump on and let any
cars waiting as well as the refill truck know that the filling process has completed. Use an Assign step to
set the GStaPumpOn to “1” and then use the Fire step, to “Fire” the EvntStartPumps event.

Figure 21.32: Tank is Full

Step 17: When the event EvntStartPumps is fired, the truck waiting for the tank to fill is activated and the
cars waiting at the pumps will be activated. If these should be activated differently, then the process needs
to be modified.

Step 18: Save and run the current model for 48 hours.

Question 7: Does this system seem to work (i.e., how many cars get no gas versus those that do)?
________________________________________________________________________________

Question 8: What happens when the tanker comes and their cars pumping gas and waiting in line?
________________________________________________________________________________

Question 9: What is the average level in the tank?

________________________________________________________________________________

Part 21.4: Reneging the Cars When the Pump Goes Off

The previous model assumed that when the tanker arrived and shut the pumps off, the cars at the pumps would
continue to wait until the pumps are turned back on. Let’s modify the model such that all of the cars waiting at
the pumps would renege (that is leave).

Step 1: Add a SINK to your model named SnkCarsRenege which will automatically generate statistics for
the reneging cars but does not needed to be connected.

Figure 21.33: Adding a new Sink that is not Connected via a Path
Step 2: When the tanker arrives, the cars waiting for a pump should be removed from the input queue of
the SrvPumpGas. After the pump is turned off, insert a Search step which can search a collection of
objects in a TABLE, a LIST, an instance of an object, a queue, as well as objects seized by the parent or
associated object. In this case, set the Collection Type to “QueueState” and the Queue State Name to
SrvPumpGas.AllocationQueue which will search the cars waiting in the input buffer to be
processed at the pump. See Figure 21.34.

356 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 · The search needs to return all the cars waiting, therefore set the Limit property to “Infinity”. The
Match Condition will allow searching for certain objects that has to be true for the item to be selected.
In this case we could have been left blank to return all entities in the queue. However, if you chose
paths to link the objects and a car was on the link and arrived at the same time the tanker arrived then
there could be a car that has seized the pump but is waiting in the “Wait” state for the pump to come
on and therefore we don’t want to remove any cars that have already seized a pump. For each car (i.e.,
EntCars) found, a new TOKEN associated with the found car is sent down the “Found” path while the
original TOKEN will eventually continue down the “Original” path once all items have been found316.

Figure 21.34: Reneging all Cars Waiting for Gas when the Pump goes Off.
Step 3: Insert a Remove step that will remove the current car associated with the “Found” TOKEN from the
SrvPumpGas.AllocationQueue. A TOKEN associated with the removed item will exit out the
“Removed” path.

Step 4: Now insert a Transfer step that will move the car (i.e., modelentity) to the SnkCarsRenege by
specifying the Node Name as Input@SnkCarsRenege as the node to send these cars. The car is
currently in the InputBuffer STATION so this step needs to transfer the entity from the current station to
the node specified as seen in Figure 21.35.

Figure 21.35: Transferring the Cars to the Renege Sink for Statistic Collection
Step 5: Save and run the model observing what happens when the Tanker arrives.

Question 10: Do the waiting cars renege (you may need to look at results)?
________________________________________________________________________________

316
Note that two tokens are now active.
Continuous Variables, Reneging, Interrupt, Debugging: A Gas Station 357

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 11: How many cars reneged in 48 hours of the simulation?
________________________________________________________________________________

Part 21.5: Interrupting the Cars When Pump Goes Off

In the previous model, the cars waiting at the pump reneged when the pump was turned off and we had already
not allowed other cars to arrive to the gas pumps while the pumps are off. The cars already in service getting
gas when the pumps go off were allowed to continue to fill up their car and then leave. Suppose now we want
the model to interrupt the cars in service and deny them further use of the pump when the tanker arrives.
These interrupted cars will leave through the SnkCarsRenege EXIT as well.

Step 1: To interrupt a process that is currently being delayed, insert an Interrupt step before the
Search step as shown in Figure 21.36 with specifications in Figure 21.37.

Figure 21.36: Adding the Interrupt to Force Cars Currently Processing to Renege
· Use the Process Name SvrPumpGas.OnEnteredProcessing which is the process that is running the
Delay step (i.e., processing of the cars). Notice that we “EndProcess” for the interrupted processes
and we establish a Limit of “4” since there are four pumps potentially in operation.

Figure 21.37: Interrupting the Cars Getting Gas

· Copy the Assign step from the step SrvPumpGas _AfterProcessing to the “Interrupted” branch
which will decrease the flow rate of each car. Then connect the Assign step to the Transfer step
that is used previously to force the cars to exit to the reneging sink (i.e., SnkCarsRenege).
· Insert a Release step that will release a Specific resource (i.e., SrvPumpGas).

Figure 21.38: Releasing the Server SrvPumpGas after Interruption

Step 2: Save and run the model observing what happens.

358 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 12: How can you tell if the interrupt is working (Hint put a break point on Search and then single
step the simulation)?
________________________________________________________________________________

Question 13: How many cars exited the renege sink now (i.e., how many cars were interrupted)?
________________________________________________________________________________

Part 21.6: Using Entities that Carry a Tank

In our gas scenario, there are really three distinct systems (i.e., the cars and the pump, the refill tanker, and
finally the continuous gas tank system). The first two systems are discrete systems that interact with the
continuous system by turning on and off the regulators as well as changing the flow rate. The [Flow Library]
provides for several other flow objects that combine both discrete and continuous together as seen in Table
21.1. Currently the tanker truck arrives and causes the FLOWSOURCE to fill up the tank at certain rate till it is
full. We assume the tanker has enough gas to fill the tank until it is full. However, the tanker truck can only
carry about 4500 gallons of gas. Now, we are going to assume that we own one tanker that has a volume
capacity (i.e., actually caries the gas) and that it will travel between the refill station and the gas station where
it fills the tank to its capacity. The tanker can move faster when it is empty (returning to the refill station).

Table 21.1: Additional Flow Objects Linking Discrete and Continuous Entities
Flow Object Description
CONTAINERENTITY A MODELENTITY that contains a TANK which is transported with the entity.
This object behaves in a similar fashion as the COMBINER. Instead of
FILLER combining a parent and member entity together, the object combines (i.e.,
fills up the tank) a discrete container entity with a flow
This object behaves in a similar fashion as the SEPARATOR does for entities
EMPTIER
by separating (i.e., emptying the tank) the flow from the container entity.
Converts discrete model entities into continuous flows (e.g., large drinking
ITEMSTOFLOWCONVERTER bottles are delivered and placed into the water machine which flows out
water to people until the bottle is empty)
Converts continuous flows into discrete entities (i.e., continuous process of
FLOWTOITEMCONVERTER spinning of fibers to produce a bobbin of yarn which is transferred to the
next station).

Step 1: Since we are going to model the truck physically transporting the gas back and forth from the refill
station, we can delete unneeded modeling components. Delete the SnkTankerLeaves as well as the
EntRefillTrucks since we need now to use a CONTAINERENTITY for the tanker.

Step 2: From the Definitions tab, insert a new EXPRESSION PROPERTY named TankerCapacity that has a
Unit Type of “Volume” with Default Units set to “Gallons” as seen in Figure 21.39.

Figure 21.39: Properties of the TankerCapacity Property

Continuous Variables, Reneging, Interrupt, Debugging: A Gas Station 359

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 3: From the [Flow Library] insert a CONTANERENTITY named CEntTankerTruck which has the
Initial Volume Capacity set to the TankerCapacity (i.e., a 4500 gallon tanker) as seen in Figure 21.40.
Change the picture to be the same tanker resizing it to be smaller. Also, you may want to make the tank
square and stretch it out.

Figure 21.40: Setting up the CEntTankerTruck

Step 4: Select the Srctanker and specify the new CEntTankerTruck as the Entity Type. Since we only
will have one tanker in the system set the Maximum Arrivals to “1” under the Stopping Conditions section.

Figure 21.41: Modifying the SrcTanker to Create One Tanker

Step 5: Next, we will create the refill station where the tanker will be filled up and then allowed to
transport the gas to the gas station as seen in Figure 21.42.

· Disconnect the FlowSrcFill from the TankGas as we will use the tanker to fill up the tank.
· Insert a new FILLER named FillTanker which allows us to fill up a container in an entity. Also, make
the processing station queue oriented to allow for the large tanker to be seen while processing.
· Connect the SrcTanker via a path to a ContainerInput node of the FillTanker.317
· Finally Connect the FlowSrcFill to the FlowInput node of the FillTanker

317The simulation will warn you when you run the simulation and the entity that is entering the container input node does not have a
tank associated with the entity (i.e., it is not a ContainerEntity).
360 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 21.42: Refill Station Area
Step 6: Once the tanker arrives at the gas station, we need to pump the gas out of the tanker and into the
tank. Insert a new EMPTIER named EmptyTanker and connect FlowOutput node to the TankGas via a
FLOWCONNECTOR. The emptier can either empty all the contents of the container entity or a specific
amount. In this case we will empty all of it until the TankGas is full (i.e., the EventStartPumps is fired
when the tank is full).

Figure 21.43: Using the Emptier to Fill up the Gas Tank from the Tanker
Step 7: Finally connect the Output@FillTanker to the Input@EmptyTanker via a time path as well as
the ContainerOutput@EmptyTanker to the ContainerInput@FillTanker. Both time paths should
have a Travel Time set to Random.Uniform(3.37,4.125) hours to travel between the gas and refill
station.

Figure 21.44: Tanker and Gas Station Refilling System

Step 8: Save and run the model.

Question 14: Are cars being allowed to get gas immediately?

________________________________________________________________________________

Question 15: How many trips does it take before cars allowed into the gas station and why?
________________________________________________________________________________

Question 16: Does the tanker always return empty?

________________________________________________________________________________

Continuous Variables, Reneging, Interrupt, Debugging: A Gas Station 361

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 17: How many cars renege during the 48 hours?
________________________________________________________________________________

Step 9: If we want the pumps to be shut off and the cars to renege like before, we will need to specify the
Before Processing add-on process trigger to be the Input_SnkTankerLeaves_Entered. Create a new
After Processing trigger that will be used to the turn the gas pumps back on as seen in Figure 21.45.

Figure 21.45: Using Process to Turn the Pumps On and Off when Filling up Tank
Step 10: If you recall the Input_SnkTankerLeaves_Entered process shuts off the pumps and then waited
for the tank to fill up before turning them back on. The EmptyTanker object is performing the filling task
now which means we need to separate the shutting down of the pumps from the starting of the pumps. For
the EmptyTanker_AfterProcessing, copy the Assign step (i.e., “PumpsOnandStopFlow”) that is after
the Wait step which turns the pumps back on. You will need to delete the assignment that shuts down the
FlowSrcFill output regulator since that is handled by the refilling station as seen in Figure 21.46.

Figure 21.46: After the Tanker Finishes Empting Turn Pumps Back On
Step 11: For the Input_SnkTankerLeaves_Entered, delete both the Assign and Wait process steps after
the Search step as seen in Figure 21.47.

Figure 21.47: Only Shutting Down the Gas Pumps and Forcing the Cars to Renege
Step 12: Save and run the model.

Question 18: How many cars renege now?

________________________________________________________________________________

362 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Part 21.7: Commentary

· Continuous variables in SIMIO can be very useful, although the rates of change (second
derivative) are limited to constants. The addition of the new flow objects has made it much easier
to use these and integrate them into discrete systems.
· For our gas station, we could have combined the gas pumps in with the continuous system as well
using the FILLER . Because the FILLER object can fill only one item (i.e., capacity of one) at a
time, you would need a FILLER for each of four gas pumps. Then, we would need to employ the
concepts from Chapter 8 to restrict flow entities to the pumps (i.e., a single queue using a
RESOURCE object). Then the cars become container entities which when created would modify
their gas tank size to specify the amount gas needed. You would connect the output of the
TankGas to the four gas pump fillers while removing the FLOWSINK.
· The uses of reneging and interruption concepts were introduced in this chapter and should not be
ignored. Reneging allows entities in queues to be removed and either destroyed or transferred.
Processes may also be interrupted. These features are not often used, but when they are needed,
they are extremely valuable. Notice also that the modeling is accomplished through processes, not
though additional specifications on the objects, as one might see in other simulation languages.
This approach grants considerable flexibility to the modeler.
· The debugging facilities in SIMIO are quite extensive. When combined with animation display
options, the trace, breakpoint, watch, and dashboard window can provide needed detail about the
behavior of the simulation.
· The value of the “Simbits” should not be underestimated. They contain numerous examples of
highly useful modeling approaches on many various topics.

Continuous Variables, Reneging, Interrupt, Debugging: A Gas Station 363

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Chapter 22
More Subclassing: Advanced Modeling of Supply Chain
Systems
Chapter 10 demonstrated the modeling of a very simple three tier supply chain system. However, only the DC
(Distribution Center) component in the supply chain used an inventory model to make decisions. Figure 22.1
shows a more complicated supply chain where two different store chains place orders on our manufacture’s
DC. The two chain DCs run inventory models similar to the manufacturer’s DC and also places replenishment
orders. The manufacture can utilize two different suppliers for a particular product. The suppliers use an
inventory model to replenish its inventory positions to satisfy demands from the manufacturer.

Reorder
Reorder Customers
Information
Information

Perfect
Walmart
Raw Material Supplier A
Product Stores/DC
Source
Flow Manufacturer Product
DC Flow
Target
Perfect Stores/DC
Raw Material Supplier B
Source

Customers

Figure 22.1: More Complicated Supply Chain

Part 22.1: Developing a Specialized Supply Chain Workstation Object

In Chapter 10, an inventory model system was built for a single inventory system. A similar approach could be
used for each segment of the supply chain in Figure 22.1 by adding additional state variables and process
logic. As the system gets more complicated, changes or the requirement of different inventory models this
approach is not very efficient. However, the ability to create specialized objects will make it easier to model
these systems with very little additional logic.

Step 1: Open up the final project from Chapter10. Now also create a new project in a separate instance of
SIMIO. It will be convenient to copy items from the previous model in constructing the new one.

Step 2: The MODELENTITY is used to model the flow of orders as well as the products. Insert the following
three discrete integer state variables into the new project model by selecting the MODELENTITY object in
the [Navigation] panel.

Table 22.1: Three New Discrete State Variables for Order ModelEntities
Variable Description
EStaOrderAmt The amount of product that is needed or has been ordered.
EStaDeliverNodeID The input node where the order should be delivered.
EStaOriginalOrderAmt The amount of the original order.

364 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 3: Our goal is to encapsulate the logic and parameters that were in the simple supply solution into an
object that can be used more than once. Select the Model from the [Navigation] panel and subclass the
WORKSTATION object318 and rename the new object SUPPLYINVWORKSTATION.

Step 4: Select the SUPPLYINVWORKSTATION object in the [Navigation] panel. Under the
Definitions→Properties, insert the same five Expression Standard Properties named InitialInventory,
ReorderPoint, OrderUptoQty, ReviewPeriod, and ReviewPeriodTimeOffset. These should all be
placed in the “Inventory” Category319 and the two review period properties need to have units of time with
default of unit of days.

Figure 22.2: Adding the Four Inventory Properties to our Supply Object
Step 5: For the SUPPLYINVWORKSTATION object, insert a DISCRETE INTEGER STATE variable named
StaInventory which will represent the current inventory amount of the product at the current segment and
another one named StaOnOrder which represents the total amount this segment has been currently
ordered from the supplier. The Initial State Value property should be zero for both the StaInventory and
StaOnOrder variables respectively.

Step 6: Next, we need to insert the same ELEMENTS as was done for the simple supply chain model into our
new object as seen in Figure 22.3 and Figure 22.4.320

· Insert/Copy two TALLY STATISTICS named TallyStatSL to track the service level performance and
TallyStatAmtOrdered to track the amount ordered.
· Insert/Copy two STATESTATISTICS named StateStatInv and StateStatOnOrder to track the inventory
level performance and the amount on order. Specify the State Variable Name property to be the
StaInventory variable and StaOnOrder respectively.
· Insert a TIMER named TimerReview to model the periodic review. The supply workstation will
review their inventory positions321 based on the reference property ReviewPeriod and start reviewing
based on the ReviewPeriodTimeOffset.

Figure 22.3: Setting up the State Statistics Track Inventory and On Order

318
See Chapter 18 for more information on subclassing objects in SIMIO.
319
The first time, you need to type the category name “Inventory” and then the next time it can be selected from the dropdown box.
320
You can copy the elements, state variables, and properties from the Chapter 10 models.
321
Remember that inventory position is the inventory “on hand” plus the inventory “on order.”
More Subclassing: Advanced Modeling of Supply Chain Systems 365

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 22.4: Setting up the Tally Statistics and Periodic Review Timer

Part 22.2: Adding the Ordering Station and Characteristics to Handle Orders

In the previous section, the same variables, properties, and elements were added to our new supply server
object that was used for simple supply chain model of Chapter 10. Recall that the model included two separate
systems (i.e., an ordering system to handle orders and one to produce the parts to fill the inventory). Figure
22.5 shows the anatomy of the new object. The original workstation will still act as the replenishment piece
(i.e., convert raw material into a finished good) but it will not send the product on to the network but update
the inventory level while the new order processing piece will ship out product based on customer orders.

Supply Chain WorkStation

Customer
Orders
Inventory Level

Replenishment via
Normal WorkStation Order Processing
Object

Supplier Ship to
Shipment Customer

Figure 22.5: Anatomy of the Supply Chain WorkStation

Step 1: Our new object will need to receive orders and process them. Therefore, a location (i.e., a
STATION) is needed to house these order entities while they are being processed. From the
Definitions→Elements→General section, insert a new STATION named OrderProcessing.322

Step 2: To set capacities of the ordering process as well as sort orders waiting, create a new referenced
property by right clicking on the property except the Parent Cost Center should be set to the
ParentCostCenter. Use the same property names as seen in Figure 22.6.

322
See Chapter 20 for more information on STATIONS.
366 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 22.6: Creating Properties to Allow User to Change the Ranking of Orders
Step 3: Switch to the Definitions→Properties panel and modify the new properties by placing the
OrderCostPerUse and OrderHoldingCostRate into the “Financials/Order” category and the rest in the
“Ordering Information” category.

Figure 22.7: All of the Properties

Step 4: To separate the processing of the ordering and production systems from the “Facility” tab, insert a
new RESOURCE named ResOrder which will be used in the ordering system while the workstation will be
used for the production system. Also, the resource should not be visible externally in the model.323

Step 5: Next, we will link the new resource capacity and ranking to the ordering station. For the Capacity
Type and Work Schedule properties create new reference properties, while for the others just choose the
ones already created.

Figure 22.8: Process Logic Properties of the New Resource

323
To make sure an object is not visible in the external view, right click the object and toggle the “Externally Visible” button.
More Subclassing: Advanced Modeling of Supply Chain Systems 367

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 6: Also, make sure to change the category of the new properties to “Ordering Information” as well as
move the two new properties up above the OrderInitialCapacity as seen in Figure 22.9. Let’s create a
Switch Condition on the new OrderWorkSchedule and OrderInitialCapacity which will not show it
unless the user selects a “WorkSchedule” as the capacity type or “Fixed” respectively. This will allow the
ordering processing system to follow work schedules. Do the same for work day and work period
exceptions.

Figure 22.9: Modifying the Order Capacity Type and OrderWorkSchedule Properties
Step 7: Next, we need to add an entry point (i.e., input node) for orders coming into the supply inventory
server as well as an exit point for reorders leaving the supply server. For the SUPPLYINVWORKSTATION
object, navigate to the Definitions→External section. Let’s add a Status Label with an expression “Sup
Inv” as on top of the picture as seen in Figure 22.10 to distinguish from the original WORKSTATION object.

ReOrder Order
Output Input

Normal Normal
Input Output

Figure 22.10: Creating the External Interface Adding Entry and Exit Nodes
Step 8: Now, move the original two “External” nodes (Input and Output) to the bottom third of the object
as seen in Figure 22.10. Insert two additional “External” nodes324 named ReorderOutput and
OrderInput. The ReorderOutput node is a TRANSFERNODE and should be right above the normal Input
node since orders will be transferred out to their supplier while the OrderInput node is a BASICNODE
placed right above the normal Output node which will accept arriving orders. This will facilitate the
notion of products flowing from the back of the chain to the front of the chain (i.e., left to right) while
orders will flow the opposite direction. The properties for both nodes are specified in Figure 22.11. Once
order entities enter the OrderInput they will be transferred directly to the OrderProcessing STATION.

324
See Chapter 17 and Chapter 20 for more information on inserting external nodes into the external view of the object.
368 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 22.11: Specifying the Properties of the ReorderOutput and OrderInput Nodes
Step 9: Next add an additional animation queue associated with the contents queue of the OrderProcessing
station as seen in Figure 22.12.325 Add the same six status labels as before which represent the inventory
level (StaInventory), average service inventory (StateStatInv.Average), and average service level
(TallyStatSL.Average).326

Processing.Contents OrderProcessing.Conten
ts

InputBuffer.Content
ss

OutputBuffer.Content
s

Figure 22.12: Adding the Animation Queues and Status Labels

Step 10: As part of the ordering system, three additional properties will be needed, as seen in Table 22.2
below. Also, the OrderProcessingTime property should have units of “Time” with the default units being
“Days” placing these properties in the “Ordering Information” category.

Table 22.2: Three Properties to Allow for the Ordering Process to Take Place
Property Type Description Required
Standard The time spent processing order
OrderProcessingTime Yes
Property→Expression before it is filled
Standard The input node id where the order
DeliverNodeID Yes
Property→Expression should be delivered
Object Reference Standard
OrderEntityType The type of order entity to send out Yes
Property→Entity
The OrderEntityType will be used in a similar fashion as the EntityType property for the SOURCE object. This
allows the object to create Reorders based on a user defined type (i.e., allows the modeler to make more
changes to the orders without having to modify the SUPPLYINVWORKSTATION object).

325
For each queue, set the Alignment to “None” under the Appearance tab.
326
Just copy the six labels from the Chapter 10 model that is currently open to get all the expressions correct.
More Subclassing: Advanced Modeling of Supply Chain Systems 369

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Part 22.3: Adding the Behavior Logic for the Ordering System

We now have defined all of the characteristics of the supply inventory object and now we need to add the
ordering system. Orders that arrive to the OrderInput node are sent directly to the OrderProcessing
STATION. At this point, the logic necessary to fill the incoming orders needs to be added. Recall from Chapter
20 on the DELAY object, STATIONS respond to entities arriving and exiting the particular station.

Step 1: Insert a new process named OnEnteredOrdering with the triggering event being
OrderProcessing.Entered as seen in Figure 22.13 to handle the processing of orders. The next
few steps will be broken down into sections in order to build the process.

Figure 22.13: The OnEnteredOrdering Process

Step 2: When entities are transferred from one location to another, the transfer has to be ended to signal to
the transferring step that is has reached its destination. Recall the OrderProcessing STATION’s capacity
was set as a reference property to allow for capacitating the order processes. Also, we added an order
processing expression that would simulate the processing of the order (i.e., checking information, credit,
etc.,) before it can be filled. We will only to try to fulfill the order if the order amount is greater than zero
and at that point we will update the service level statistics. Figure 22.14 shows the first few steps of the
process while Figure 22.18 shows the properties associated with the process steps.

Figure 22.14: First Five Steps of Handling the Arriving of Orders

· Insert an End Transfer step to end the transfer of the arriving order entity from the
OrderInput external node into the OrderProcessing STATION.
· Insert a Seize step that will seize one unit of capacity from the “Specific” resource which is the
ResOrder as shown in Figure 22.15.

Figure 22.15: Seizing a Unit of ResOrder Capacity

· Like the normal WORKSTATION we would like to provide the capability for the user to execute add-
on process triggers before processing the order and after the order has been processed, insert two

370 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Execute steps. Create a new reference property named OrderBeforeProcessingAddOn and
OrderAfterProcessingAddOn as the Process Name property as seen in Figure 22.16. In the
Definitions→Property section, change the category of this new reference property to “Add on
Process Trigger” and the display name to “Order Processing” and “Order Processed” respectively.

Figure 22.16: Order Processing and Processed Add-on Process Trigger

· Next, insert a Delay step between the two Execute statements to represent the processing of
the order before filling the order using the existing OrderProcessingTime property.

Figure 22.17: Setting up the Delay Step

· Add a Decide step to determine if the order amount is greater than zero before continuing on to
the fulfillment process.
· Update the service level statistic using a Tally step using the same logic from .

Figure 22.18: Properties Associates with Beginning of the Order Process

Step 3: The previous steps were similar to the ones used in the simple supply chain model in Chapter 10.
These were the only ones associated with the ordering process since the DC was the object of interest and
orders did not flow to the downstream customer in the supply chain. The actual filling and shipping of the
order to the customer will be handled by the last steps of the process as seen in Figure 22.19.

Steps Copied from

PerformTeardownActivity

Figure 22.19: Steps Used to Fill the Order and Ship it out to the Next Element
Step 4: The first section (i.e., the first three steps) will update the inventory position by the amount being
shipped out. If the current inventory position cannot handle the order amount (i.e., Decide step) then set
More Subclassing: Advanced Modeling of Supply Chain Systems 371

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 the order amount equal to the current inventory level.327 If the inventory is large enough (i.e., “False”
branch) go directly to decrementing the current inventory position by the order amount. Insert the
Decide and the two Assign steps with the properties in Figure 22.20. Otherwise, we need to change the
order amount of the entity to let the customer know how much we were able to ship to them.

Figure 22.20: Updating the Inventory Position

Step 5: After the inventory position has been updated, the shipment has to be sent out the normal Output
external node. If the deliver node “id” was specified (i.e., ModelEntity.EStaDeliverNodeID >0)
in the MODELENTITY then we will set the destination node before transferring the shipment out. Otherwise
just ship it out, and let the links handle where the shipment will go (see Figure 22.21).

Figure 22.21: Shipping the Order Out to Downstream Customer

Step 6: If the output buffer capacity is greater than zero, we will transfer it to the output buffer, otherwise
the order is sent out automatically. As shown in Figure 22.19, copy the Decide, Assign and two
Transfer steps from the end of the “PerformTeardownActivity” process since it is the same (see Figure
22.28).

Step 7: Next, link the “False” branch from the Decide on whether they specified a deliver node to the
Decide of the output buffer greater than zero.

Step 8: Insert a new process named “OnExitedOrderingProcess” with the triggering event being
OrderProcessing.Exited as seen in Figure 22.22 which it executes the “ExitedAddOn” process
trigger if it is specified. Copy all three steps from the “OnExitedProcessing” process. This process will
run and execute all the steps when the output buffer capacity is zero and the orders are directly sent to the
external node. Modify the Release step to release the “Specific” resource ResOrder rather than the
capacity of the parent object, changing the “Units per Object” to 1.

327
Before we just set the inventory level to 0 if the order amount was greater than zero but we are now shipping it out.
372 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 22.22: Process that Executes when Entities leave the OrderProcessing STATION.
Step 9: Since the orders are being shipped out through the output buffer, we need to override the
“OnEnteredOutputBuffer” so it releases the ResOrder rather than the parent object like the previous step.
Click on the process and override and modify the Release step to release one unit of ResOrder.

Figure 22.23: Modifying the OutputBuffer Entered Process

Part 22.4: Adding the Behavior Logic for the Inventory Replenishment System

We now have defined the logic for the ordering and fulfillment system. Next, the logic necessary to implement
the inventory replenishment system will be added to the new object.

Step 1: Setup the periodic review by creating a process to respond to the timer TimerReview. From the
“Processes” tab of the SUPPLYINVWORKSTATION object, create a new process named InventoryReview.
Set the Triggering Event property to respond to the TimerReview.Event which causes the process to
be executed every time the timer event fires, as seen in Figure 22.24. This is almost identical to the same
process created in Chapter 10. Copy the entire process and all steps can be just copied and then the
Create, Assign and Transfer steps need to be modified. You will need to modify all the GSta variables to
Sta since they are part of the workstation.

Figure 22.24: Process to Reorder

· The Decide step does a “ConditionBased” check to see if the current inventory is less than
the reorder point (i.e., StaInventory < ReorderPoint )
· If a new reorder is needed, use the Create step to create a new entity as seen in Figure 22.25
which creates the entity based on the OrderEntityType reference property this time.

More Subclassing: Advanced Modeling of Supply Chain Systems 373

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 22.25: Creating a New Reorder to Send the Supplier
· For the newly created order, use an Assign step to assign the order amount and then increase
the WIP value based on the order amount. Additionally, we will also set the original amount
to help model reliability of the suppliers as well as the deliver node, if it has been specified, as
in Table 22.3.
Table 22.3: Determining the Order Amount and Updating the WIP
State Variable New Value
Math.Max(0,OrderUpToQty –
ModelEntity.EStaOrderAmt
StaInventory–StaOnOrder)
StaOnOrder StaOnOrder +
ModelEntity.EStaOrderAmt
ModelEntity.EStaDeliverNodeID DeliverNodeID (This is a property)
ModelEntity.EStaOriginalOrderAmt ModelEntity.EStaOrderAmt
· Insert an Execute step that will run a user defined add-on process trigger which will be
executed before the reorder exits the system to be sent to the supplier (see Figure 22.26). In
the Definitions→Property section, change the category of this new reference property to
“Add-On Process Triggers” and the display name to “ReOrder Exited.”

Figure 22.26: Inserting a User-Defined Add-on Process Trigger

· Once the OrderEntityType has been set, it needs to be sent to the supplier by using a
Transfer step to move it from “FreeSpace” to the ParentExternalNode of the
ReOrderOutput.328

Figure 22.27: Sending the New Reorder to the Supplier

Step 2: Recall from Chapter 10 when the products arrived back at the DC and had been packaged, the
StaInventory variable was increased by the order amount, while the StaOnOrder variable was reduced
by the value in the “Processed” add-on process trigger for the WrkDC. Since we are encapsulating this
same logic inside our new SUPPLYINVWORKSTATION object, the same logic needs to be added right after
the “Finished Good Operation” trigger executed in the PerformTeardownActivity process. This process is
run when the batch of parts has finished processing and the tear down activity has finished for the batch.

328
Now the modeler will indicate where the reorders should travel via a link from the ReOrderOutput node.
374 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Select the PerformTeardownActivity process from the “Processes” tab and then click the override button
( ) which allows you to modify it as seen in Figure 22.28.329

Executes Finished Good

Operation Add-On Trigger

Figure 22.28: The Default OnEnteredProcessing Process

Step 3: After the execution of the “Finished Good Operation” add-on process trigger as seen in Figure
22.28, the entity will be transferred to the output buffer if the buffer capacity is greater than zero (i.e.,
“True” branch) or it will execute any before exiting assignments and then transferred to the external output
node. In our make-to-stock model, the product is placed into inventory and not sent on to the next
process. Therefore, you need delete all the steps after the Execute step and then insert three new
process steps (i.e., Assign, Release, and Destroy) as seen in Figure 22.29.

New Steps Added

Connect to Release
Step

Figure 22.29: The New PerformTeardownActivity Process

· The Assign step updates the Inventory and WIP values as specified in Table 22.4.

Table 22.4: Updating Inventory and DC once Product Arrives

State Variable New Value
StaInventory StaInventory +
ModelEntity.EStaOrderAmt
StaOnOrder StaOnOrder -
ModelEntity.EStaOriginalOrderAmt

· Since the entity is no longer sent to the Output Buffer (i.e., it will not enter the OnEnteredOuputBuffer
process) or to the external node which will not run the ExitedProcessing process, the workstation
capacity is never released. Therefore, copy the Release step from the OnExitedProcessing process
after the Assign step and make sure Exclusion Property is empty (i.e. reset).
· Insert a Destroy step to destroy the entity order that has arrived since it will not continue.

329
See Chapter 19 and Chapter 20 for more information on overriding and restoring processes.
More Subclassing: Advanced Modeling of Supply Chain Systems 375

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 · Make sure to connect the end of the FinishedBadOperation add on process execution step to the
Release step.

Recall, the initial inventory property was used to set the beginning StaInventory position in the
OnRunInitialized process. Override the OnRunInitialized process of our new object and insert the same
Assign step as before as seen in
Step 4: Figure 22.30.

Figure 22.30: Setting the Beginning Inventory Position

Part 22.5: Using the New to Model the Complex Supply System

The new SUPPYINVWORKSTATION object has been defined and can now be used to build the model (see Figure
22.31) of the complex supply chain system of Figure 22.1. 330

Figure 22.31: SIMIO Model of the Complex System

Step 1: In the model “Facility”, add a new MODELENTITY named EntOrders. Add a status label that is
connected to the entity in similar to the one in Chapter 10 with the Expression property set to
EStaOrderAmt as seen in the top right of Figure 22.31.

330
You should spread out the objects until the path direction arrows show. Having paths and connectors in the wrong direction can
create some “hard to find” errors. When you are sure the paths and connectors are in the proper direction, place the objects back into
their appropriate places.
376 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 2: Insert two SOURCES named SrcWalmartCust and SrcTargetCust that will represent the daily
number of customers arriving at the stores to purchase the products (interarrival times are one each day).
As was done before, set the Before Exiting State Assignments that will set the
ModelEntity.EStaOrderAmt to Random.Poisson(70) and Random.Poisson(50)in days,
for the Walmart and Target stores respectively to represent the number of daily orders for each customer
type. Here is a case where if we had done the assignment in a tokenized process it would have been
easier.

Step 3: Next add a new SUPPYINVWORKSTATION named SupInvWalmart that will represent the inventory
position object of Walmart.

· Set the Processing Batch Size to two, Operation Quantity to ModelEntity.StaOrderAmt and
the Processing Time property to Random.Triangular(1,2,3) in minutes to reflect the
processing of the product into the stores and on the shelves as seen in Figure 22.32.

Figure 22.32: Setting up the Processing of the Batches

· Under the Inventory property category, set the Review Period to every five days, the Initial Inventory
to 1100, the Reorder Point to 900 with the Order Up to Qty to 1100 as seen in Figure 22.33.
· For the Order Information, it takes a ¼ of day to process an incoming order. Also, set the Deliver
Node ID to [email protected] which will allow orders sent to the supplier to be sent
back to the Walmart store since two different stores are ordering on the same supplier.

Figure 22.33: Inventory and Ordering Information Properties of the Walmart Store
Step 4: Repeat Step 3 for the Target store by copying the Walmart SupInvWalmart naming the object
SupInvTarget changing the properties seen in Figure 22.34 making sure to change the Deliver Node ID to
[email protected] and the inventory properties since Target does not service as many
customers.

331
Recall each object in SIMIO has a unique identification number which is accessed by the ID function.
More Subclassing: Advanced Modeling of Supply Chain Systems 377

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 v
Figure 22.34: Inventory and Ordering Information Properties of the Target Store
Step 5: Connect the two sources to their appropriate OrderInput node at the SupInvWalmart and
SupInvTarget via connectors as seen in Figure 22.31.

Step 6: Insert two SINKS into the model named SnkWalmart and SnkTarget. Connect the Output nodes
of the SupInvWalmart and the SupInvTarget SUPPYINVWORKSTATIONS to the respective sinks via
CONNECTORS.

Step 7: Next we need to create the distribution center which will service the two stores. Copy the
SupInvWalmart object and name the new SUPPYINVWORKSTATION object SupInvDC to represent the
DC.

· Set the Processing Time property to the following expression in minutes.

Random.Triangular(5,10,20)
· Set the Initial Inventory property to 1800, the Reorder Point property to 1400 and the Order
Up to Qty property to 1800.
· Set the Delivery Node ID to 0.

Step 8: Next we need to connect the DC to each of the two stores.

· For products to flow to the stores, connect the Output node of SupInvDC as seen in the
whole model of Figure 22.31 to the Input nodes of SupInvWalmart and SupInvTarget via
TIMEPATHS. It takes generally four days to transport product from the DC to each store.
However, it can take a minimum of three or a maximum of seven days to travel from the DC
to each store. Therefore set the Travel Time property to Random.Pert(3,4,7)days.
· For orders to flow from the stores to the DC, connect the ReorderOutput node of both stores
to the OrderInput node of the SupInvDC via connectors to represent EDI (Electronic Data
Information) order transmissions.

Question 1: With our new object, how can we model the situation where Walmart orders are satisfied
before Target orders, if they arrive at the same time?
________________________________________________________________________________

Step 9: In the first portion, we have utilized one supplier to supply the DC, which is a make-to-stock type
supplier. Therefore, copy the SupInvDC and paste in a new SUPPYINVWORKSTATION object SupInvA as
seen in Figure 22.35. Only a few things will actually change.

· Set the Processing Time property to Random.Triangular(10,20,30) (minutes) since

it takes longer to convert the raw material into a product.
· Set the Initial Inventory property to 600, the Reorder Point property to 300 and the Order Up
to Qty property to 600.
· Set the Delivery Node ID to 0.

378 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 22.35: Connecting a Supplier to the DC.
Step 10: Similar to the stores, connect the ReorderOutput node of the SupInvDC to the OrderInput node
of the SupInvA via a connector to have orders flow back to the supplier.

Step 11: Like the DC, products take the same amount of time to ship from the Supplier A to the DC.
Therefore, connect the Output node of SupInvA to the Input node of the SupInvDC via TIMEPATH with
the same Travel Time property set to Random.Pert(3,4,7)days.

Step 12: Supplier A has a perfect raw material supplier (i.e., we assume the material is always available).
Therefore, use a connector to connect the ReorderOutput of SupInvA directly to the Input node of
SupInvA which allows the supplier to order a particular amount of product from its perfect raw material
supplier to replenish its inventory position through its own manufacturing.

Step 13: Save and run the model for 52 weeks observing the output.

Question 2: What are the average service levels of each of the segments in the supply chain?
________________________________________________________________________________

Question 3: What is the current and average inventory level of each segment in the chain?
________________________________________________________________________________

Question 4: Do you think the current system is adequate and what seems to be the problem?
________________________________________________________________________________

Part 22.6: Adding a Secondary Supplier for Overflow Orders

By reviewing the results, most of the inventory positions were zero which led to very low service levels. The
central issue is the lead time length and variability of getting product from the back of the chain to the front of
the chain. There are many ways this situation can be fixed by optimizing reorder points and order up to
quantities. Also process improvement projects might reduce the processing times or shipping times. Let’s add
a second supplier (Supplier B) which is a make-to-order operation to handle orders from the DC as well. In
many situations multiple sourcing strategies exist and can be modeled.

Step 1: Insert a new WORKSTATION named WrkSupB to act as a make-to-order supplier (i.e., when orders
arrive they are produced and not stored ahead of time). Modify the model as shown in Figure 22.36.

More Subclassing: Advanced Modeling of Supply Chain Systems 379

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 22.36: Add Make-to-Order Supplier B
· Connect the ReorderOutput of SupInvDC to the Input of the WrkSupB via a CONNECTOR link to
allow orders to be sent to the new supplier electronically.
· To allow products to flow, connect the Output node of the WrkSupB to the Input node of the
SupInvDC via a TIMEPATH. It takes Random.Pert(1,2,3) days to ship then product to the DC
from this supplier.
· Set the Processing Time property of the WrkSupB to produce the products to
Random.Triangular(8,10,20) (minutes). Again set the Operation Quantity to
ModelEntity.EStaOrderAmt and the Processing Batch Size property to two.

Step 2: In this example, the DC has access to the Suppliers A current inventory level. If the order amount is
greater than the current inventory level, the DC will order only up to the inventory position and then order
the remaining portion from Supplier B. Insert a new “ReOrderExited” add-on process trigger that is
executed when the order exits the WrkSupB as seen in Figure 22.37.

Figure 22.37: ReOrder Exit Add-on Process Trigger

· Insert a Decide step that checks to see if the ModelEntity.EStaOrderAmt is less than or equal
to the SupInvA.Inventory position. If it is (i.e., the “True” branch) then use a Set Node step to
specify that the order should be sent to the Supplier A (i.e., OrderInput@SupInvA).

380 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 22.38: The Properties of the Decide and Set Node for Reordering Process
· If the order amount is bigger than the current inventory position then decide if supplier A has any
inventory at all (i.e., SupInvA.StaInventory >0). If it does not (i.e., “False” branch) then send
the current order to supplier B via the Set Node with Node Name property set to
Input@WrkSupB. Otherwise, Create a copy of the current order and set the “Original” order
amount to the current inventory level as seen in Figure 22.39.

Figure 22.39: Create Step to Copy the Current Order

· In the assignment, as shown in Figure 22.40 set the inventory position at SupInvA to the
StaOrderAmt and the original order’s StaOriginalOrderAmt needs to be set equal to the new
StaOrderAmt. Then go ahead and send the order on to the SupInvA by moving the “End” point on
top of the Set Node as seen in Figure 22.37.

Figure 22.40: Set Order Amount and the Original Order Amount
· For the newly created order which is a copy, set the StaOrderAmt to the remaining portion and also
update the StaOriginalOrderAmt as shown in Figure 22.41.

Figure 22.41: Creating an Order to Send to the Second Supplier

Step 3: Then transfer from “Free Space” to Input@WrkSupB as seen in Figure 22.42.

Figure 22.42: Transfer Input at WrkSupB

Step 4: Save and run the model for 52 weeks observing the output.

More Subclassing: Advanced Modeling of Supply Chain Systems 381

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 5: What are the average service levels of each of the segments in the supply chain?
________________________________________________________________________________

Question 6: What is the current and average inventory level of each segment in the chain?
________________________________________________________________________________

Part 22.7: Commentary

· This model provides a framework for many interesting embellishments and much added realism.
· Failure information could be linked with the ResOrder resource. The ability to suspend the
ordering process when failures occur should also be added.
· Cost information can easily be added to the orders and objects using the cost parameters under the
financials and one can choose suppliers as well as have different shipping options (i.e., air, rail,
and trucking) to assist in meeting consumer demand to make it even more realistic.
· The reliability of the supplier can be added to the new SUPPLYINVWORKSTATION object such that
shipping or picking errors can be modeled (suppose a customer asks for 100 but only 98 are
shipped, even though there is sufficient inventory).
· The object can be modified to handle multiple products rather than a single product type.
· Additional inventory policies can be built into the model that would allow the user to choose the
particular inventory policy for each segment of the supply chain.
· We could also add a MONITOR event to create a continuous review system rather than just a
periodic review system. Users could choose between the systems.
· The WORKSTATION was chosen as the object to subclass rather than the SERVER since it can process
a batch of parts correctly. However, you cannot increase the capacity of WORKSTATIONS like you
can SERVERS by changing the capacity. You can change the batch size but that is not quite the same
as the average flow time for the parts would be correct but the variance would not. One would
have to insert multiple workstations to achieve the same thing.

382 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Chapter 23
More Subclassing: Process Planning/Project Management
Part 23.1: Process Planning

A new product (called Product 3) is being introduced. One unit of Product 3 is produced by assembling one
unit of Product 1 and one unit of Product 2. Before production begins on either Product 1 or Product 2, raw
materials must be purchased and workers must be trained. Before Products 1 and 2 can be assembled into
Product 3, the finished Product 2 must be inspected. The table below gives the list of activities and their
predecessors and the average duration of each activity.

Table 23.1: Activity Precedence and Duration

Activity Predecessors Duration (days)

A = train workers -- 6
B = purchase raw materials -- 9
C = produce Product 1 A, B 8
D = produce Product 2 A, B 7
E = test Product 2 D 10
F = assemble Products 1 and 2 C, E 12

The typical project management diagram for the example can be seen in Figure 23.1.

Figure 23.1: Sample Project Planning Example

Clearly the project will take an average of 38 days and activities B, D, E, and F are critical (“on the critical
path”) – meaning if any of these increases, then the time to completion is increased. But in reality, activity
completion times are uncertain (random) and different activities may become critical. So we can’t rely on
deterministic values for the activity times and instead need to represent the times with random variables.

Also, activities might share resources to complete the task. Therefore, the goal will be to determine the critical
path and the probability that a path will be on the critical path.

Table 23.2: Project Activity Statistics

Earliest Finish The time the current task finishes (i.e., entity leaves the path).
The time the current task could finish without changing the critical path
Latest Finish
or causing the successive tasks to start later.
Slack The time difference between the Earliest Finish and the Latest Finish
Percent CP The percentage of the time the current path is on the critical path.

Earlier chapters have demonstrated the ability to specialize/subclass existing objects which allows additional
characteristics and behaviors to be added or existing ones to be modified, but none have specialized one of the
More Subclassing: Process Planning/Project Management 383

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 link objects (i.e., PATHS or TIMEPATHS). We will model the project management problem using the
TIMEPATHS to consume time. We will also use one replication to represent a single completion of the project,
so multiple replications will be used to statistically characterize the project.

Part 23.2: Creating a Specialized TIMEPATH to Handle Activities

The ability to create new objects is one of the advantages of SIMIO and the TIMEPATH link is an attractive
object to use for the activity time, since the entities can be seen visually moving along the paths. The paths
represent the important objects (i.e., critical path determination). Each path needs to keep track of its own
statistics (i.e., Earliest Finish, Slack, Percent CP, etc.). The path also needs to know when an entity reaches
the end of the path, and also restrict only one entity to travel on each path. This encapsulation of the time path
gives the object the “intelligence” to automatically calculate the statistics, as well as allow us to utilize it for
many different project management problems.

Step 1: Create a new model and subclass the TIMEPATH and change the Model Name property to
PRJTIMEPATH332.

Step 2: Select the PRJTIMEPATH from the [Navigation] panel. Insert two DISCRETE REAL STATE variables
named StaEarliestFinish and StaCalculateSlack. The first state variable will be used to keep track of the
time the activity finishes while the second one will be used to signal when to calculate the slack.

Step 3: To keep track of critical path calculations, add three TALLY statistics named
TallyStatEarliestFinish, TallyStatSlack, and TallyStatPercentCP via the Definitions→Elements
section.

· Change the “Category” property to “Critical Path” for all three TALLY statistics under the “Results
Classification.
· Change the Data Item property to “Earliest Finish”, “Slack”, and “Percent CP” for the appropriate
statistic.
· Also, for the TallyStatEarliestFinish and TallyStatSlack, set the Unit Type to “Time.”

Step 4: Recall, only one project will be completed for each replication. In order to model a “junction” from
which multiple activities start, the paths should not allow more than one traveler and travel should be in
one direction only. From the Definitions→Properties, click on the down arrow to reveal the inherited
properties.

· Select the IntialTravelerCapacity property and change the Default Value from Infinity to one and the
Visible property to “False” so a user cannot change it and it will automatically be set.
· Select the InitialDesiredDirection property and change the Default Value to forward and the Visible
property to “False” so a user cannot change it and it will automatically be set.

332
Right click on the object in the [Navigation] section and select properties.
384 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 23.2: Setting the Default Values of the Travel Capacity and Desired Direction
Step 5: The TIMEPATH defines seven processes that determine the behavior of the link as seen in Table
23.3.

Table 23.3: Processes of the TimePath

Process Description
Process runs when the leading edge of an Entity, Worker, or Transporter
OnEntered
reaches the beginning of the TIMEPATH.
Process runs when the trailing edge of an Entity, Worker, or Transporter leaves
OnExited
the TIMEPATH.
Process runs when the leading edge of an Entity, Worker, or Transporter
OnReachedEnd
reaches the end of the TIMEPATH.
OnRunEnding Process runs once at the end of the simulation.
OnRunInitialized Process runs once at the beginning of the simulation.
Process runs when the trailing edge of an Entity, Worker, or Transporter enters
OnTrailingEdgeEntered
the path.
Step 6: Once the project ENTITY reaches the end of the path, the activity has finished and the earliest finish
time can be calculated. Override the “OnReachedEnd” process as seen in Figure 23.3 to perform the
calculation.

New Steps Added

Figure 23.3: Calculating Earliest Finish and Shutting Down Path

· Insert an Assign step that will set the DesiredDirection to
Enum.TrafficDirection.None which will shut down the path and not allow anymore entities
to travel this path. Also, set the StaEarliestFinish state variable to the current time (Run.TimeNow).
Refer to Figure 23.4.

More Subclassing: Process Planning/Project Management 385

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 23.4: Shut down the Path and Set the Earliest Finish
· Insert a Tally step to add the one observation of the TallyEarliestFinish statistic. as seen in Figure
23.5.

Figure 23.5: Add an Observation of the Earliest Finish Statistic

Part 23.3: Creating a Junction Object to Handle Precedent Constraints

In the previous section, we modeled the project activities as specialized TIMEPATHS which automatically
calculated all the statistics. However, they cannot enforce the precedence constraints or start up activities nor
determine when the latest finish happens. The precedence constraints have to be enforced by the junctions
which also initiate the next set of activities. Therefore, the junction has to wait until all the preceding activities
have finished before starting all of the succeeding activities at the same time. Also, the junctions determine
the latest finish of all the preceding activities based on the time the last of the preceding activities finishes.

Step 1: None of the standard objects can model this circumstance easily and most standard objects are too
complicated to modify for this purpose. A new object similar to the DELAY object of Chapter 20 will be
created. From the “Project Home” tab, insert a new “Fixed Class Model” named JUNCTION which will be
used to model the junctions.

Step 2: We need our junction to act as a start node and an end node, as well as intermediate junctions
within the project network. Start nodes will start activities at the start of the project while the end nodes
will terminate all activities at the end of the project.

· Insert a new String LIST from the Definitions→List→Create section named ListJunctionType with
three entries as seen below. This will allow the user to choose the Junction Type.

Figure 23.6: Defining Different List Types

386 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 · Insert a new standard “List Property” named JunctionType which allow the modeler to specify the
junction type with “Junction” as the default type and place it in the “Project Management” category as
seen in Figure 23.7.333

Figure 23.7: Specifying the List JunctionType Property

Step 3: Like the standard SOURCE, if the JUNCTION object is a “Start Node” it will need to create entities to
be sent out into the project network. From the Definitions→Properties→Add→Object Reference¸ add an
Entity property named EntityType making sure to place it into the “Project Management” category
created in the step before. Also, you only want the property displayed if the JunctionType property is set
to “StartNode”.

Figure 23.8: Specifying the Start Node

333
Note, that you can change the caption of the list object as a space was added between “StartNode” and “EndNode” options.
More Subclassing: Process Planning/Project Management 387

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 4: Each junction needs to know the number of precedent activities to wait on and the number of
successive activities to start. Insert two new DISCRETE INTEGER STATE properties named
StaPredecessors and StaSuccessors which will be set at the beginning of the simulation run.334

Step 5: Insert a DISCRETE INTEGER STATE variable named StaActivitiesCompleted which will keep
track of the number of activities that have completed and a DISCRETE REAL STATE variable named
StaLatestFinish that will record when the time the last activity has completed.

Step 6: As was done in Chapter 20, the DELAY object, the incoming entities need a location where actions
can occur. From the Definitions→Elements insert a new STATION named Project with all the default
values.

Step 7: Next, let’s define the external view seen by the user of the JUNCTION object.

· From the Definitions→External→Drawing section, use a Polygon shape to draw a red diamond as
seen in Figure 23.9.
· Insert an EXTERNAL NODE that is a BASICNODE on the left edge of the red diamond that will accept
incoming activities and automatically send them to the Project STATION.
· Next create the output of activities by inserting an External Node that is a TRANSFERNODE on the right
edge of the red diamond that will route successor activities.

Figure 23.9: External View of the Junction Object Along with Input and Output Nodes
Now that we defined the external view and activities (i.e., entities) can flow in and out of our object, we need
to add the process logic to handle the precedence and the starting of the activities as well as determining the
number predecessors and successors.

Step 8: From the “Process” tab, selected the OnRunInitialized process from the Select Process drop
down. Insert an Assign step that will set the number of predecessors (i.e., StaPredecessors to
Input.InboundLinks.NumberItems) and the number of successors (i.e, StaSuccessors to
Output.OutboundLinks.NumberItems).

334In Versions prior to 5 there was no way to access the number of links coming or going out of BASICNODES or TRANSFERNODES and
these values had to be passed in as a number.
388 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 23.10: Initializing the Number of Successors and Predcessors
Step 9: Create a new process named “OnEnteringProject” that will execute every time an activity (i.e.,
ENTITY) enters the Project station by specifying the Triggering Event to be “Project.Entered” as seen in
Figure 23.11. The process needs to update the number of activities finished. If the number of activities
equals the number of predecessors, it will create the number of successors and send them out to the
network otherwise there are still activities that need to complete and we will destroy the current activity
entity.

Figure 23.11: Process to Handle Activities that Enter the Junction

· As always, the first thing you need to do is to end the transfer of the entity object into the station by
inserting an End Transfer step.
· Next update the number of activities that have completed by one.

Figure 23.12: Updating the Number of Activities by One

· Now decide if all the precedent activities have completed (i.e., StaActivitiesCompleted ==
StaPredecessors) as seen in Figure 23.13.

More Subclassing: Process Planning/Project Management 389

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 23.13: Deciding if All Preceding Activities are done
· Note the use of the Exclusion Expression property which can be used at the start of the simulation to
determine if this step is necessary.335 If the JUNCTION is a “StartNode” or only has one preceding
activity then the next set of activities can be automatically started (it will be set to one) or this is an
“EndNode” then these entities can be automatically destroyed (i.e., set to two). If you know at the start
of the simulation, skipping steps can greatly speed up the simulation since it does not need to be
evaluated. Set Exclusion property to the following.

((StaPredecessors==1 &&JunctionType != List.ListJunctionType.EndNode)

|| JunctionType == List.ListJunctionType.StartNode) + 2*(JunctionType
== List.ListJunctionType.EndNode)

· If the total number preceding activities has not been completed (i.e., still waiting on some activities to
complete), then insert a Destroy step to kill the current activity entity.
· If this activity is the last one to complete or the junction is a “Start Node”, then we need to start the
next set of successor activities. Insert a Create step that will create copies of the current Entity (i.e.,
set the Create Type property to “CopyAssociatedObject”). The Number of Objects will be set to
StaSuccessors -1 since one activity ENTITY is already present. Note the use of the Exclusion
Expression property to skip this step if there is only one successor as seen in Figure 23.14.

Figure 23.14: Creating the Successor Activities

· Using Transfer steps, the original activity ENTITY needs to be sent from the “Current Station” to
the “ParentExternalNode” Output while the newly created activity entities need to be sent from
“Free Space” to the same ParentExternalNode Output as seen in Figure 23.15.

Figure 23.15: Sending Successor Activities Out to the Network

335
If the expression evaluates to a one then then the step will not be included and the entity will flow directly out of the true or primary
branch. If it evaluates to a two then it would be skipped and flow directly out of the false or secondary branch. All other evaluations
will have the process step evaluated each time.
390 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 10: If the JUNCTION is a “Start Node” then it must automatically create the start activity to start this
branch of the network. Therefore, the “OnRunInitialized” process which is executed for each object at the
beginning of the simulation will be used to handle this logic. Select the “OnRunInitialized” and add a
Create and Transfer steps after the Assign sets the number predecessors and successors.

Figure 23.16: “OnRunInitialized” Process to Start Activities

· Insert a Create step that will produce one new object of EntityType property. Again, right click on
the Object Instance Name property and select the referenced property. Again, note the use of the
Exclusion Expression property to skip this step if it is not a “Start Node”. Refer to Figure 23.17.
· If it is a start node and an activity has been created, then insert a Transfer step to move it from
“FreeSpace” to the Project Station.336 This will then run the “OnEnteringProject” process which will
produce the correct number of successors without having to repeat the code as seen in Figure 23.17.

Figure 23.17: Create and Transfer Steps for the “OnRunInitialized” Process
Step 11: A warning may appear stating that there is nowhere for the entity to go once it exits the last
JUNCTION. SIMIO now warns you when entities are sent to nowhere and will be automatically destroyed.
This warning helps modelers make sure this was no modeling error. In this case, this is not an error but not
a good idea to still have the warning. Therefore, if this is an “End Node” JUNCTION we do not want to
transfer the entity out of the Output ParentExternalNode.

· Select the Transfer step connected to the “Original” path and add 2*(JunctionType ==
List.ListJunctionType.EndNode) to the Exclusion Expression property which will force
the model entity out of the “Failed” branch. A value of two sends the entity automatically out the
failed.
· Next, drag the “Failed” branch to the Destroy step so entity will be destroyed now.

336
Note you could have sent the entity from “FreeSpace” to the Input ParentExternalNode as well. Either way it runs the process.
More Subclassing: Process Planning/Project Management 391

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 23.18: Updated Model that will Destroy the Entities if the Junction is an End Node

Part 23.4: Creating Small Network to Test the New Object

At this point our junction object seems to be setup to handle the precedence constraints and start new
activities. Go back to the original model and create a very simple network to test this ability. At this point we
will not worry about naming the objects.

Step 1: Insert a new MODEL ENTITY named EntActivities and four Junction objects leaving their names as
the default. Place them in the configuration as seen in Figure 23.19.

String.format("The successor or predecessor {0} connected to {1} is not a PrjTimePaths. Change the link to
PrjTimePath",Link.Name, Name)

Figure 23.19: Simple Network to Test the JUNCTION Object

Step 1: Connect Junction1 to Junction2 via two PRJTIMEPATHS with one path taking one day and the
other one taking two days and then connect Junction2 to Junction3 and Junction4 via one day
PRJTIMEPATHS.

Step 2: Set Junction1 to be a “Start Node” while Junction2 should be just a “Junction” as seen in Figure
23.20.

392 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 23.20: Setting up the Start Junction and the Secondary Junction
Step 3: Save and run the model observing what happens.

Question 1: Did two activities get created from Junction1 and Junction2?
________________________________________________________________________________

Question 2: Did Junction2 wait before both activities were completed before starting the next set of
activities?
________________________________________________________________________________

Step 4: It looks like only one was created when actually both were placed on the same path. Click on the
Output node of Junction2 and change the Outbound Link Preference from “Any” to “Available” which is
used to select the next link. In this case we only want to select from the ones that are currently available.

Step 5: Save and run the model observing what happens.

Question 3: Did two activities get started from Junction2 now?

________________________________________________________________________________

Step 6: At this point, the user will have to change the default property from “Any” to “Available” every
time which is not acceptable since they may forget. Therefore, we will automatically set it for the modeler
by specializing the TRANSFERNODE object. Subclass the TRANSFERNODE object and name it
PrjTransferNode.

Step 7: Select the PRJTRANSFERNODE from the [Navigation Panel]. From the Definitions→Properties,
click on the “Properties (Inherited)” down arrow to show all the inherited properties. Select the
OutboundLinkPreference property and change the Default Value to “Available” and set the Visible
property to “False” so the user cant accidently set it back.

Step 8: Return back to the JUNCTION object and select the external Output node in the
Definitions→External section. Change the Node Class property from the TRANSFEROBJECT to the new
PRJTRANSFERNODE as seen in Figure 23.21.

Figure 23.21: Changing the Output node to be PrjTransferNode

More Subclassing: Process Planning/Project Management 393

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Part 23.5: Building the Example Network

Let’s build the simple example from the beginning section.

Step 1: Rearrange the current JUNCTIONS in the model and insert a new JUNCTION using the default names
as seen in Figure 23.22.

Figure 23.22: Example Project Management Model

Step 2: Use PRJTIMEPATHS to represent each of the activities by connecting two JUNCTIONS together.
Table 23.4 should be utilized specify the name of each of the paths as well as to set the Travel Time and
Units properties.
Table 23.4: Activity Times and Names
From To Name Travel Time Units
Junction 1 Junction 2 A Random.Pert(5,6,10) Days
Junction 1 Junction 2 B Random.Pert(6,9,9) Days
Junction 3 Junction 4 C Random.Pert(6,8,16) Days
Junction 2 Junction 3 D Random.Pert(4,5,7) Days
Junction 2 Junction 4 E Random.Pert(6,7,10) Days
Junction 4 Junction 5 F Random.Pert(8,12,16) Days
Step 3: Next, use the values in Figure 23.23 to specify the Project Management properties of each of the
five junctions to make sure the precedent constraints are satisfied and the number successor activities are
started.

Figure 23.23: Properties of the Five Junctions

Step 4: Save and run the model, make sure the run length is “Unspecified” which will stop the simulation
when no more entities are in the system.

394 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 4: How many projects were completed and what was the total completion time?
________________________________________________________________________________

Step 5: Remember, we decided to model one project completion with one replication. Therefore, insert a
new experiment that will replicate the project management network 1000 times.

Step 6: Run the experiment and look at the results under the “Pivot Grid tab. Change the Unit Settings of
time to Days.

Step 7: Filter the “Category” column by clicking on the funnel in the right hand corner (see Figure 23.24)
and select only the “Critical Path” statistics and then filter “Statistic” column to only include the
“Average” since we only have one observation per replication. Figure 23.25 shows the reduced statistics
where the project will take between 27 and 38 days with the average taking 33 days plus or minus 1.8
days337.

Figure 23.24: Filter the Category only to Include Critical Path Statistics

Figure 23.25: Earliest Finish Statistics for 1000 Replications

Part 23.6: Adding the Slack and Percent of time on Critical Path Calculations

At this point we can determine the Earliest Finish (EF) time of all the paths and the total time to complete the
project, but it is difficult to determine what paths constitute the critical path since path times are random.
However, the slack time for each path and the percent of time this path is on the critical path can assist the
project manager. The Earliest Finish (EF) time can be calculated without any other information from other
objects. However, the computation of the Latest Finish (LF) requires knowing the slack and the number of
times the path is on the critical path.

The Latest Finish time is only known at the junctions and is determined when the last activity finishes. When
two SIMIO objects need to communicate with one another, there are two ways to approach this
communication. In the first approach shown in Figure 23.26, the JUNCTIONS have to know the PRJTIMEPATHS
that are connected to them. The PRJTIMEPATHS will monitor a state variable within their object that is set to
one by the JUNCTION when the last activity finishes. In the second approach the PRJTIMEPATHS need to know
the JUNCTIONS that they are connected too and they will monitor a state variable inside the JUNCTION object.

337
Your numbers may be a little different since the random numbers may be sampled differently in your model, due to the way the
modeling components were added. However the results should be statistically identical.
More Subclassing: Process Planning/Project Management 395

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 We will illustrate the first approach in this section where the monitoring state variable resides inside the
monitoring.

PrjTimePath PrjTimePath
Junction
Monitor
Monitor
Signals/Sets

Calculate Slack Signals/Sets

Junction Calculate Slack
State Variable State Variable

Figure 23.26: Approaches to Signaling from One Object to Another

Step 1: Save the current model as a Chapter 23-6.spfx because we will need to come back to the previous
section model to create the second approach so make sure it is saved before renaming it.

Step 2: Select the PRJTIMEPATH object from the [Navigation] Panel.

Step 3: From the Defintions→Elements tab, insert a new MONITOR named MonitorSlack which will be
used to calculate the slack calculation when the StaCalculateSlack state variable changes values to one as
seen in Figure 23.27.

· Set the Monitor Type to “DiscreteStateChange” and the State Variable Name to the
StaCalculateSlack variable.
· Create a new Triggered Process Name to execute when the monitor event changes.

Figure 23.27: Setting up the Monitor to Calculate the Slack

Step 4: Once the last activity has happened, the associated JUNCTION will set the StaCalculateSlack state
variable to one. Since this variable is being watched by the MonitorSlack, it will execute the “Triggered
Process” as seen in Figure 23.28.

· Insert a Tally step that will calculate the slack (LF-EF) where the latest finish is the current time and
add either a zero or one to the TallyPercentCP to indicate whether or not the path is on the critical
path or not. You need to select the Tallies (More) repeating group to set both statistics.

Figure 23.28: Calculating the Slack and Determining if it is on Critical Path

396 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 5: The PRJTIMEPATH object is now setup to calculate the two new statistics when its
StaCalculateSlack variable is changed. The JUNCTION object must now set the StaCalculateSlack
variable for all of the precedent paths once the last activity has finished. Select the JUNCTION object from
the [Navigation] panel.

Step 6: After the last precedent activity has finished (JUNCTION object’s OnEnteringProject process), the
StaCalculateSlack variable will need to be set to one for all the precedent paths. Insert a Search step
that will use to search for the correct activity paths right after the Decide step. Recall the search step
returns a TOKEN associated with the found object (i.e., PRJTIMEPATH) which can then be used to set its
own variable.

Figure 23.29: Setting the StaCalculateSlack Variable of the Paths When All Activities have Finished
· Figure 23.30 shows the Search step that will search a “NodeInBoundLinks” where the Node Name
property is set to Input which is the parent input node. Note, this node will not be in the dropdown list
but will need to be just typed into the field. The Exclusion Expression (i.e., JunctionType
==List.ListJunctionType.StartNode) specifies to skip this Search step if this JUNCTION
is a “StartNode” and does not need to notify in predecessor links.

Figure 23.30: Searching for all Inbound Links

· Insert an Assign on the “Found” branch and specified with the values in Figure 23.29 which will set
the PrjTimePath.StaCalculateSlack discrete variable to one.

More Subclassing: Process Planning/Project Management 397

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure 23.31: Setting the Calculate Slack Variable
Step 7: Save and rerun the experiment.338

Question 5: What percentage of the time is path A on the critical path?

________________________________________________________________________________

Question 6: How long could we delay the start of activity A and not have any effect on the critical path?
________________________________________________________________________________

Question 7: How long could we delay the start of activity C and not have any effect on the critical path?
________________________________________________________________________________

Step 8: Looking at the results it is clear that the critical path is B D E F since these paths have the highest
percentages of time on the critical path as seen in Figure 23.32 for Percent CP average.

Figure 23.32: Results of the Critical Path Analysis

Part 23.7: Adding Slack and Percent of time on CP Calculations Second Approach

The previous section uses the JUNCTION objects to signal (i.e., set the StaSlackVariable) the activity time
paths of the Latest Finish. In this section, we will take the alternative approach and let the PRJTIMEPATH
monitor a JUNCTION state variable. This section demonstrates how an object can monitor state variables of
other objects, which can prove to be very useful in communicating among objects.

Step 1: Save the current project (Chap23-6p2.spx) under a new name.

338
Sometimes when adding properties to objects that are already in the model causes ghost errors of things not being specified. Save
and reopen the model to eliminate them if this occurs.
398 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 2: Remove the the Search steps from the “OnEnteringProject” process and insert a new Assign
step that assigns the current time (Run.TimeNow) to the StaLatestFinish state variable.

Figure 23.33: Setting the StaLatestFinish State Variable

Step 3: Next, select the PRJTIMEPATH object.

Step 4: From the Definitions→Properties→Add→Standard Property drop down, insert a new STATE
property named JunctionLF which will allow the modeler to specify the JUNCTION state variable to
monitor. Change the Category Name to “Project Management” by specifying it.

Figure 23.34: Create a State Variable Property

Step 5: Modify the MONITOR element MonitorSlack to monitor the new state reference property
JunctionLF as seen in Figure 23.35.

Figure 23.35: Using the Reference State Property to Monitor

Step 6: Return back to the model and for each PRJTIMEPATH, specify the appropriate JUNCTION’S
StaLatestFinish state variable.

· For example (see Figure 23.36), Junction LF State Variable property would be set to
Junction2.StaLatestFinish for both paths A and B.
· For paths C and E, D, and F it should be set to Junction4.StaLatestFinish,
Junction3.StaLatestFinish, and Junction5.StaLatestFinish respectively.

Figure 23.36: Specifying the State Variable

Step 7: Save the model and rerun the experiment comparing the results of the two systems.
More Subclassing: Process Planning/Project Management 399

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Question 8: Did you get the same results?
________________________________________________________________________________

Step 8: The problem was the last path did not calculate the latest finish statistics since it was heading to an
“End” junction. Select the JUNCTION object and then inside the OnEnteringProject process. Recall that all
the entities were destroyed via the exclusion property of the Decide step (see Figure 23.13) and therefore
it is not sent to the Assign step. Therefore, remove the portion of the exclusion property associated with
end node: 2*(JunctionType == List.ListJunctionType.EndNode).

Step 9: Save the model and rerun just the model.

Question 9: Did anything happen when the activity went through the end node?
________________________________________________________________________________

Part 23.8: Error Checking to Make Sure Modeler Uses Junction Correctly

There are several assumptions the model has made. We are assuming the modeler will always connect up the
junctions correctly using only the new PRJTIMEPATH links. To eliminate this assumption, we will modify the
OnRunInitialized to check for these issues, warn the modeler, and then end the simulation run as seen in
Figure 23.37.

Figure 23.37: Checking for Valid Links

Step 1: Insert a new STRING STATE variable named StaWarningMessage to specify the particular
message.

Step 2: Insert a conditional based Decide step to check to see if there are predecessor links connected to
the Junction. If this is not a “Start Node,” the step will be skipped automatically as specified by the
Exclusion Expression property. Insert another Decide step that will check to make sure the successor
links have been defined if this is not an “End Node” as seen in Figure 23.38.

Figure 23.38: Checking to see if the Predecessors and Successors have been Defined

400 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 3: If the Predecessors>0 Decide steps fail (i.e., not start node but links are connected to the
junction), insert an Assign step on the failed branch that will set the StaWarningMessage state variable
to a string which states the JUNCTION name as well as what to do fix the issue.339

String.format("{0} is Not a Start Node and there are no predecessors (i.e,

Project Path links) defined", Name)

Step 4: Next, if the Successors>0 Decide steps fail, insert an Assign step on the failed branch that will
set the StaWarningMessage state variable to a string which states the JUNCTION name as well as what to
do fix the issue.

String.format(" {0} is Not an End Node and there are no successors (i.e, links)
defined",Name)

Step 5: Connect both of the Assign steps to the same Notify step. The Notify step can either write a
message in the Trace window or give a pop-up Warning to the user. We will use the StaWarningMessage
as the Message Content property as seen in Figure 23.39.

Figure 23.39: Sending a Warning to the User

Step 6: Insert an End Run step that will terminate the simulation since there is an error in the model and
we don’t want to continue.

Step 7: The first two decide steps determined if there were no predecessor or successor links defined.
However, the user could connect the JUNCTIONS using normal PATHS, CONNECTORS, or TIMEPATHS which
is also an error. Insert two Search steps that will search both the inbound and outbound links as seen in
Figure 23.40. The inbound search will be skipped if this is a “Start Node” while the outbound search will
be skipped if the JUNCTION is an “End Node” owing to the Exclusion Expression property.

Figure 23.40: Searching to Make Sure the Predecessor and Successor Links are PrjTimePaths

339 Note the use of String.Format function to create string that has one argument (i.e., “{0}”) which is set to the value of the
JUNCTION NAME.
More Subclassing: Process Planning/Project Management 401

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Step 8: When links are found, we need to check to make sure they are PRJTIMEPATH links only. Connect
both “Found” branches to a conditional Decide step that checks to see if these are PRJTIMEPATH using
the IS keyword which returns False if these objects are not project time paths.

Figure 23.41: Checking to see if the Links are Project Time Paths
Step 9: If any of the links are not project time paths, then set the warning message which will tell the
modeler which link is not a project time path that is connected to which JUNCTION as seen below.

String.format("The successor or predecessor {0} connected to {1} is not a

PrjTimePaths. Change the link to PrjTimePath",Link.Name, Name)

Step 10: Next, connect the Assign step to the Notify step by dragging the End on top of the step.

Step 11: Select the Facility window and test the new error checking by inserting another Junction into the
model.

Step 12: Save and then run the model with the new Junction not connected.

Question 10: Does the simulation stop and warn the user with the Junction in error?
________________________________________________________________________________

Step 13: Next connect both the input and ouput to other nodes via paths and rerun the model.

Question 11: Does the simulation stop and warn the user with the Junction in error?
________________________________________________________________________________

Part 23.9: Commentary

· Sometimes the activities within a project employ common resources. The resources can limit the
number of simultaneous activities, and resource allocation and the timing of the allocation become
limits on the project progress. The use of common resources can be incorporated directly into the
PRJTIMEPATHS. However deciding which activities should be given priority for the resources can
become a problem.

· Another possible option in project management is determining the amount of time to delay the start of
an activity. By delaying the start of an activity, resources can be made available to more critical ones
and thus the project completion reduced. Optquest could be used to optimally determine the start time.

402 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Appendix A
Input Modeling
Any discussion of creating a “realistic” model of any operation eventually turns to the issue of variability or
randomness. Incorporating “randomness” into your simulation model can be a significant challenge because
the causes of randomness are difficult to understand completely. As a result we approximate this randomness
with some kind of “randomness model”, usually a statistical distribution. Because we must “approximate” the
randomness, we view this as “modeling”. And because the randomness is not explained by the simulation
model, this randomness must be “input” to the simulation. Thus reason for the term “Input Modeling.” There
are several forms of input models for randomness.

One input model is the time-based arrival process, which in SIMIO is called the “Time Varying Arrival Rate”
arrival mode at a SOURCE. Fundamentally, we are trying to model an arrival process that changes throughout
time. The simplest model is a Poisson arrival process whose mean changes with time, which is generally
referred to as a Non-Homogeneous Poisson Process (NHPP). Recall that the Poisson has only one parameter,
namely its arrival rate (mean). A NHPP has a Poisson mean with an arrival rate change that can be any
function of time. In SIMIO the rate function is a stepped function which provides an arrival rate during a
specified time period. In SIMIO, the rate is always given in events per hour, while the time periods have a
fixed interval size (and number). A more sophisticated NHPP would permit a more flexible rate function.

Another model of randomness is the specification of a statistical distribution for some observed randomness,
usually an amount of time. The most prominent examples are Processing Times at SERVERS, WORKSTATIONS,
COMBINERS, and SEPARATORS and Travel Time in a TIMEPATH. Other examples of times include Transfer-In
Time, Setup Time, and Teardown Time in various objects. Also the “Calendar Time Based” failure type at
working objects requires uptime between failures and time to repair. Creating input models for this kind of
input may employ some way to find a statistical distribution to match the randomness behavior. Sometimes
there is randomness in a number, such as the size of an arriving batch or the number of parts in a tote pan.

Part A.1 Random Variables

Randomness is thus represented in simulation by random variables. These random variables describe either a
continuous value (such as time) or a discrete value (such as a quantity). While you may not know how the
randomness occurs, you might know something about the distribution of these values such as some descriptive
statistics. Descriptive statistics include measures of central tendency such as a mean, median, or mode.
Maybe it’s a measure of variability such as a variance, standard deviation, coefficient of variation, or range.
Other descriptive statistics include the measure of asymmetry called the skewness and the measure of the
peakedness called the kurtosis.

Part A.2 Collecting Data

More than likely you don’t know much and will need to collect some data from the real input process to get
some idea of how the randomness behaves. This data is simply a “sample” from which you might want to
estimate how the population is distributed.

However before we begin to examine how to use “sample data”, you need to remember the most important
rule of data collection: Build the simulation model before collecting data!

The process of collecting data is almost always an onerous, frustrating, and expensive. You need to know that
you absolutely need the input model before you mount a data collection effort. The best way to know that you
need particular data is to construct the simulation model first. The simulation model doesn’t know if the input
models were obtained from real data. Simply use your best judgment to obtain input models.

More Subclassing: Process Planning/Project Management 403

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 One suggestion is to interview some people familiar with the process that is generating the randomness and try
to elicit from them some general characteristics. If you can only estimate a minimum and a maximum, use a
uniform distribution to match. However, generally, you can obtain three estimates: a mode, a minimum, and a
maximum. For example, suppose we want to estimate the randomness that describes the amount of time it
takes you to get to work or school. If I say, tell me the most likely or most common time it takes and you say
15 minutes. That estimate of 15 minutes becomes a mode. Now tell me how long it takes if you make all the
lights and cruise here without any difficulty, namely the minimum amount of time. Suppose you say 10
minutes. Next give me the maximum, when everything goes wrong. You say the worst is 30 minutes. With
these there estimates use a PERT distribution (sometimes called the BetaPERT) as shown in Figure A.1.

Figure A.1: Pert Distribution

As we you see, this distribution insures we bound our randomness to realistic values. In this case there are no
negative values340, the distribution is usually skewed to the right, and the mode is less than the mean. We
highly recommend this means of obtaining estimates, especially in the early modeling stages. The Triangular
distribution can be used also to model the case when given the minimum, maximum, and most likely point.
However, it will be heavier in the tails while the Pert will have more probability around the mode341.

With the liberal use of these kinds of estimates, build your simulation model. Vary some of the input model
parameters (i.e., the mode, minimum or maximum) and see what impact they have on the simulation outputs.
In particular, note how they impact the simulation outputs that are important to you. With a little
experimentation of this type – called a “sensitivity analysis”, you begin to get a sense of which data are most
critical. You can marshal a data collection effort that is proportional to the importance of the input model.
There are a number of hard-to-answer questions regarding data collection. Often you can’t collect as much
data as you would like, so you will need to consider the limitations of your data collection budget. You will
want to collect as much data as you can, bearing in mind that not all data is equally important in the model.

Part A.3 Input Modeling: Selecting a Distribution to Fit to Your Data

Once you begin the data collection effort, you can begin to select distributions that fit your data. Why do we
want to select a distribution to fit the data and not simply use the data directly in the simulation model? First,
the data is only a sample from a population and not a total population, so we don’t know exactly whether the
sample completely represents the population from which it is sampled. Secondly, real data tends to be

340 If a sample from a distribution for a time, like a processing time, produces a negative value, SIMIO will display an error.

Distributions that can have negative values, like the Normal must be used with care. You can avoid negative values using and
expression like Math.Max(0.0, Random.Normal(10.0,4.0).
341
The PERT distribution does have some concerns. Its variance changes significantly as the mode approaches the end points. Some
have suggested that it should not be used if the mode is within 13% of the end points.
404 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 incomplete in that it doesn’t completely match the process being modeled. Third, we believe that there are
enough standard statistical distributions from which to choose that will do a good job of modeling the “true”
input process. So we will look for a distribution to fit the data.

Fortunately, once we have some data, the process of finding a distribution to fit the data is somewhat routine.
Here are the steps we will follow:

1) Collect data and put into spreadsheet

2) Identify candidate distributions
3) For each candidate distribution
a) Estimate parameters for the candidate distributions
b) Compare candidate distribution to histogram
c) Compare the candidate distribution function to the data-based distribution function
4) Select the best representation of the input (the input model)

Candidate Distributions

Before going into much detail, it is useful to do a SIMIO “Help” on “distributions”. There will you see
descriptions of all the distributions that SIMIO supports. Presently they include the standard “continuous”
distributions of Beta, Erlang, Exponential, Gamma, JohnsonSB, JohnsonSU, LogLogistic, LogNormal,
Normal, PearsonVI, Pert, Triangular, Uniform, and Weibull as well as the standard discrete distributions of
Binomial, Geometric, NegativeBinomial, and Poisson. Also there are the empirical distributions called
Continuous and Discrete, which represent a data-based continuous and discrete distributions. You should
spend some time looking through these distributions so you have some sense of their “shape” and parameters.
For instance, Figure A.2 displays the Erlang from the SIMIO help.

Figure A.2: Erlang Distribution

From the display, you see that the Erlang is a non-negative distribution specified by a mean and a number of
phases (each phase being Exponentially distributed with a mean of the distribution divided by the number of
phases).

In the prior figure, there is the display of a histogram. A histogram is simply a display of the frequency that
data values fall within a given range. Typically the ranges or cells of a histogram have identical ranges
(sometimes the end cells are longer) and the number of histogram cells range from 5 to 15. The histogram is
an arbitrary display since the user determines the size of the cells and the number of cells. There are some
“rules of thumb” used. For example, by dividing the range of observations (maximum – minimum) by the
number of cells, you can determine the range for each cell. Although you would like to keep the number of
cells, C, between 5 and 15, the following can be a good guide:

More Subclassing: Process Planning/Project Management 405

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 5 13
Scott’s Rule: C @ (n )
3
Surges’Rule: C @ 1 + 3.3log10 (n)

Here the number of observations in your data is n. With the range for each cell, you simply display the portion
of the observations that fall within each range. By examining the histogram, you may want to select certain
distributions to fit. But before you fit a distribution to the data you need the parameters for the distribution
and they will need to be estimated from the data.

Estimating Distribution Parameters

The estimation of distribution parameters employ one of two methods: moment match and maximum
likelihood. Moment matching requires that you know, for a two parameter distribution, the mean and variance
of the distribution in terms of its parameters. If the Erlang mean is given by the symbol θ and there are k
phases to the Erlang, then the

Mean = θ

Variance = θ2/k

Therefore if you compute the average and the sample variance from your data as and s2 , approximate mean

by the average and the variable by the sample variance. Solving, you can approximate the parameters of the
Erlang by the following equations

θ=

2
k=(

The method of “matching” the sample statistics to the parameter values is the essence of moment matching.

Maximum likelihood requires more statistical development than we wish to devote here. However the basic
idea is to find the parameters of the distribution so they have the greatest “likelihood” of coming from that
parameterized distribution. Most any statistical book that deals with “estimation” will contain a description of
maximum likelihood. In actual practice, the process of finding the maximum likelihood can become a
computational challenge, if done by hand, so using some kind of software to perform the computations is
highly desirable. We do need to note that most statisticians prefer the use of maximum likelihood to moment
matching for parameter estimation. The reasons are somewhat technical, but the in general maximum
likelihood will have the best parameter estimates (they are optimal in the sense they produce the best unbiased,
minimum variance estimates).

Determining Goodness-of-Fit

Once a candidate distribution has been identified and its parameters determine, you can now compare the
distribution you fit to the data. Here is an example of a histogram with a candidate distribution (Lognormal)
being displayed in Figure A.3. The distribution parameters have been computed from the data.

406 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure A.3: Probability Density Function
The basic question is how good is the “fit” of the Lognormal to this data? Our answer is based on two
approaches: visual inspection and statistical tests.

Visual Inspection: Visual inspection is the most important strategy because it visualizes the fit. Looking at
the prior histogram and the fitted distribution, it’s most important to see how the fitted distribution matches the
observed histogram particularly in the tails. The “tails” are most important because the variation in the
distribution is largest in the tails and it is the tails the most generally affects the queuing, utilization, and flow
time statistics.

There are several other ways to “see” how well the fitted distribution matches the data. Most are based on the
cumulative distribution function (CDF). For example, the comparison of the fitted CDF versus the data-based
CDF is given in Figure A.4.

Figure A.4: Cumulative Distribution Function

Again, we are looked how well the fitted distribution represents the data. When looking at the difference, it is
useful to plot the difference in the two CDFs as in Figure A.5.

More Subclassing: Process Planning/Project Management 407

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure A.5: Difference between Empirical CDF and Fitted CDF
In this plot we are looking to see if the difference is generally above or below 0, since that would indicate the
fit overestimates or underestimates the data.

Two of the more useful visual plots is based on the “distance” between the fitted CDF and the data-based
CDF. If you measure the difference vertically, you obtain the PP plot, which emphasizes the difference in the
center of the distribution. The PP plot is shown in Figure A.6.

Figure A.6: P-P Plot Example

If you measure it horizontally, you obtain the QQ plot, which emphasizes the difference in the tails of the
distribution. The QQ plot is shown in Figure A.7.

Figure A.7: Q-Q Plot Example

408 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 In both the PP and QQ plot, we want to see the points “close” to the dividing line. Deviations from the line
indicate a potential for overestimation or underestimation of the data. Finally, we tend to have more interest in
the QQ plot than in the PP plot since it emphasizes the difference in the tails of the distribution, where we
believe the fit is most important.

Statistical Tests: Statistical tests are based on classical statistical testing in which a null hypothesis is
compared to an alternative. The null hypothesis is that the data comes from the fitted distribution. But the
“significance” of a statistical test is when the null hypothesis is rejected (Type 1 error). But in goodness-of-fit,
we want to accept the null hypothesis so the power of the test is relevant (Type 2 error) and hence we would
like are larger p-value. Thus statistical tests tend to only reject the poorest of fitted distributions. It is the
reason we place so much emphasis on the visual inspection.

There are three standard goodness-of-fit statistical tests: the Chi-Squared test, the Kolmogorov-Smirnov test,
and the Anderson Darling test. These tests are more thoroughly described in statistical texts. However we can
provide a general description of each.

The Chi-Squared test is based on comparing the observed number of observations in a histogram cell with the
number expected if the fitted distribution was the input model. Although, the test is often based on equal cell
widths, it is generally better to use equal probability cells for the test.

The Kolmogorov-Smirnov (K-S) is a non-parametric test the looks at the largest (vertical) distance between
the fitted CDF and the data-based CDF. The test is only approximate if parameters have to be estimated
(which they usually are). A generalization of the K-S test is the Anderson Darling which puts more weight in
the tails of the difference (the K-S uses equal weights).

Part A.4 Distribution Selection Hierarchy

The following is a rule of thumb and represents the distribution selection hierarchy one should follow.

1. Use recent data to fit the distribution using an input modeling software
2. Use raw data and load discrete points into a custom distribution (i.e., Empirical CDF)
3. Use the distribution suggested by the nature of the process or underlying physics
4. Assume a simple distribution and apply reasonable limits when lacking data (e.g.,Pert or Triangular)

It is always better to fit distributions to data collected about the process. If no good fits can be achieved the
empirical distributions can be used to model the data. If there is some underling distribution suggested about
the process owing scientific processes then one can utilize these. If you have very little or no data, then
assuming a simple distribution and applying reasonable limits can be used as a start. Often we can obtain the
minimum, maximum and sometimes the most likely. In this case we can use the Pert distribution.

Part A.5 Empirical Distributions in SIMIO

In contrast to the standard distributions, for which a functional form must be determined parameters must be
estimated, empirical distributions are based totally on your data. You have two choices for empirical
distributions: discrete and continuous. These distributions are represented in Figure A.8 which describes the
probability mass/density function as well as the cumulative distribution function.

The SIMIO statements for the distributions in Figure A.8 are specified according to the following cumulative
distributions:

Random.Discrete(v1, c1, v2, c2, …,vn,cn)

Random.Continuous(v1, c1, v2, c2, …,vn,cn)

More Subclassing: Process Planning/Project Management 409

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Discrete Continuous

Figure A.8: Empirical Distributions in SIMIO

Of the two distributions the Discrete is the most often used. We might have a case where 30% have a value of
2.0, 50% have a value of 3.1, and 20% have a value of 6. A plot of this case is shown in Figure A.9.

Sum of probabilities = 1.0

Figure A.9: Random.Discrete(2, 0.3, 3.1, 0.8, 6, 1.0)

Some other examples of the use of the empirical discrete distribution in SIMIO is:
Random.Discrete(2, 0.3, Math.Epsilon +5.2, 0.8, 6, 1.0)

Random.Discrete(Random.Exponential(0.5,0.3,Run.TimeNow,0.8,6, 1.0)

Random.Discrete(“Red”, 0.3, “Green”, 0.8, “Blue”, 1.0)

410 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Of course the value/expression can be negative, although it occurs infrequently. Recall also the the cumulative
probability parameters/expression must be in the range 0 to 1 and the last cumulative probability parameter
must be 1.

An example of the use of continuous distribution might be where 30%of the time the value falls (uniformly)
between 1 and 2.5, 60% between 2.5 and 4, and finally 10% between 7 and 8. The plot for this case is shown
in Figure A.10.

Sum of probabilities = 1.0

Figure A.10: SIMIO Specification: Random.Continuous(1,0,2.5,0.3,4,0.9,7,0.9,8 1.0)

The values/expression should not decrease, although SIMIO makes no check. Also cumulative probability
parameters/expression values must be in range 0 to 1 and the last cumulative probability parameter must be 1.

Part A.6 Software for Input Modeling

The computational aspects of the distribution finding process have been incorporated into software. A number
of software “packages” exist now to help in fitting a distribution to data (SAS Jmp, ExpertFit, EasyFit,
Minitab). The software will do the work of finding candidate distributions that represent the data.

Typical of input modeling software is EasyFit from http://www.mathwave.com. It comes with a free trial
version and a variety of licenses may be obtained. The software has a catalog of over 50 standard
distributions. Data can be read from worksheets or entered and edited in the software’s worksheet. The fitting
system considers all the distributions as possible candidates. The user, may however, limit the fitting to
(continuous) distributions which have an upper or lower bound or both and whether these bounds are open or
closed. Maximum likelihood estimates are used for distribution parameter estimation for each candidate
distribution. The visual displays include the probability density (or mass) function, the cumulative distribution
function, the PP and QQ plots, and the CDF difference plot. The histogram of the data can also have the
number of bins (cells) specified. The statistical goodness-of-fit tests include the K-S, the Anderson-Darling,
and the Chi-Squared with equal probability or equal width cells. The candidate distributions can be ranking by
the statistics produced from the statistical tests.

Whenever you use input modeling software, you need to be concerned with parameter correspondence
between the software parameters and the SIMIO parameters. For example, the EasyFit parameters of
Erlang (m,β) correspond to Random.Erlang(m*β,m) in SIMIO. Table A.1 shows how to convert EasyFit
parameters to SIMIO parameters

More Subclassing: Process Planning/Project Management 411

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Table A.1: Easy Fit to SIMIO Conversion of Input Parameters
Distribution Easy Fit SIMIO
Bernoulli Bernoulli(p) Random.Bernoulli(p)
Beta Beta(α,β) Random.Beta(α,β)
Binomial Binomial(nt,p) Random.Binomial(p,n)
Erlang* Erlang(m,β) Random.Erlang(m*β,m)
Exponential* Exponential(λ) Random.Exponential(λ)
Gamma* Gamma(α,β) Random.Gamma(α,β)
Geometric Geometric(p) Random.Geometric(p)
Johnson SB** JohnsonSB(γ,δ,λ,ξ) Random.JohnsonSB(γ,δ,ξ,ξ+λ)
Johnson UB(SU)** JohnsonSU(γ,δ,λ,ξ) Random.JohnsonUB(γ,δ,ξ,λ)
Log-Logistic* Log-Logistic(α,β) Random.LogLogistic(α,β)
Lognormal* Lognormal(σ,μ) Random.Lognormal(μ,σ)
NegativeBinomial Neg.Binomial(n,p) Random.NegativeBinomial(p,n)
Normal Normal(stDev,mean) Random.Normal(Mean,stDev)
PearsonVI* Pearson6(α1,α2,β) Random.PearsonVI(α1,α2,β)
Pert Pert(m,a,b) Random.Pert(a,m,b)
Poisson Poisson(λ) Random.Poisson(λ)
Triangular Triangular(m,a,b) Random.Triangular(a,m,b)
Uniform Uniform(a,b) Random.Uniform(a,b)
Weibull* Weibull(α,β) Random.Weibull(α,β)
*---Distributions that have a "Location" parameter, but was not included in the syntax. For an example
on how to use a location parameter in a SIMIO random variable distribution, see below.

**---Distributions whose "Location" Parameters were indicated in the syntax, to show how to derive
some of the parameters in SIMIO of the same distribution.

For many distributions EasyFit generates a “location” parameters along with the distribution parameters. This
location parameter permits a distribution to be located at some arbitrary point. If EasyFit provides a location,
say δ, for the Erlang then you will form the SIMIO expression:

δ + Random.Erlang(m*β,m)

The location parameter is used to shift a distribution so it will not be based at zero.

Part A.7 The Lognormal Distribution

The Lognormal Distribution may require some explanation in case you want to use it. The Lognormal is
defined as the distribution you get by taking the logarithm of a normally distributed random variable. That
normal distribution defined with a mean µ and a standard deviation σ are, in fact, the parameters of the SIMIO
and the EasyFit Lognormal distribution. However they are not the mean and standard deviation of the
Lognormal distribution which is given by:

2
/2
LogMean = em+s
2 2
LogStd = e 2m+s (es - 1)

412 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 If you have data from which you compute x and sx , then you can compute SIMIO parameters (the Normal
distribution parameters) from:

m » ln( x 2 / x 2 + s x 2
s » ln(1 + s x 2 / x 2 )

Part A.8 Modeling the Sum of N Independent Random Variables

We are often faced with determining the processing time for a batch of parts to be processed individually. The
processing time is, therefore, a sum of N independent random variables. For example, if there were 25 parts in
a batch and each one took a Triangular(a, m, b) minutes to be processed, then the total time to process the
batch is given by the summation of 25 independent triangular distributions as seen in the following equation.

Often people will mistakenly model the processing time of summation of N independent random variables as
N times the random variable of interest which is incorrect as it does not correctly create the appropriate
distribution, but it is easy to specify.

To illustrate the point, let’s take the Triangular distribution given a minimum value of a, mode of m, and a
maximum value of b. The mean and variance of this distribution is given in the following table.

Mean

Variance

Because it is much easier to just perform the multiplication rather than the summation, people will utilize the
multiplication form which is given by the following equation.

As one can see in Table A.2 below the means of the two different versions are the same however the variances
are not. In other words the modeling distribution is not what is desired. Figure A.11shows the PDF of each of
the distribution and one can see that the true underlying distribution (i.e., summation distribution) is quite
different. The reason is the variances of the summation of N independent variables add together rather than
the standard deviations.

More Subclassing: Process Planning/Project Management 413

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Table A.2: Calculation of the Mean and Variance of Summation and Multiplication Methods

Summation
Mean

Summation
Variance

Multiplicat
ion Mean

Multiplicat
ion
Variance

Summation Distribution

Multiplication
Distribution

Figure A.11: Comparing True Summation Distribution with the Multiplication Distribution

Correct Handling of Summation of N Independent Variables

The most appropriate method is to sample N random variables and then sum these values to reach the correct
sample. The WORKSTATION will model this exact scenario correctly since each order is made up multiple items
and a processing time was given based an item. Figure A.12 shows a simplified version of the Processing
process that handles the summation through batching. As you can see the Delay step will sample one random
variable and then delay the associated entity. The End Activity step checks to see if there are any more
batches. If there are more to perform, the TOKEN will loop back and delay the entity again. The Delay step
will be invoked N times sampling an independent random variable each time. For example, the Operation
Quantity property would be set to 25 for example with the Processing Batch Size set to one. Therefore, 25
separate batches or delays will be performed.

414 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Figure A.12: Simplified Batch Processing similar to the Workstation

Approximation for the Summation of N Independent Variables

As N gets large, the distribution that results from the summation of N independent distributions becomes
Normal owing to the Central Limit Theorem. Therefore, given a large enough N, the summation distribution
can be approximated using a Normal distribution with the mean and standard deviation (STD) in Table A.3.
Figure A.13 shows the exact summation distribution, the normal approximation and the incorrect
multiplication distribution. As you can see the approximation is very good given N was 25.

Table A.3: Calculation of the Mean and Variance of Summation and Multiplication Methods

Normal Mean

Normal STD

Summation Distribution
Normal Approximation
Distribution

Multiplication
Distribution

Figure A.13: Simplified Normal Approximation versus the True Summation Distribution
The disadvantages of using the normal approximation is the sampled values can become negative and the
value N needs to be sufficiently large enough. Depending on the distribution, N < 10 will not yield a very good
normal approximation and typically N larger than 20 is sufficient.

More Subclassing: Process Planning/Project Management 415

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Index
.Node.AssociatedStationLoad, 116
A .Node.NumberRoutingIn, 83
Add Additional Symbol .Resource.Failure, 208
ModelEntity, 79, 85, 136, 238, 297 .Server.Inputbuffer.Contents, 83
Animated People, 29 .Server.Processing.Contents, 83
Parts, 65 .Worker.ID, 242
People, 246 Definition, 242
Vet Patients, 210 Dynamic Selection Rule, 236
Resource, 291, 303, 346 Search Match Condition, 223, 252
Add-On Process Triggers. See Process Add-On Process Changeovers, 139
Triggers Combiner, 79
Double Clicking to Create, 94 Match Members, 81
Allocation Secondary Resources, 87
Queue, 225, 278, 279, 356 Confidence Interval
Allocation Queue Equation, 36
Animated, 304 Connector, 57
Animated Queue, 129 ContainerEntity, 359
Parking Station, 190 Conveyor, 199, 290
Animated Queues, 28 Accumulating, 199
Alignment Options, 28 Entity Alignment, 200
Animation, 27 Non-Accumulating, 199
Circular Gauge, 347 Stopping and Starting, 201
Status Labels, 151, 158, 346, 369 Costing. See Financials
ModelEntity, 220, 316, 377 Create Symbol, 343
String Formatting, 161
Status Pie, 140
Status Plot, 89, 346
D
Animiated Queue Data Table. See Tables Data Table
Alignment, 123 Debugging
Appointment Schedules, 124 Break Point Button, 240
Arrival Events Per Time Slot, 127 Detached Queue, 123, See Animated Queue
Arrival No-Show Probability, 127 Double clicking
Arrival Time Deviation, 127 Process Trigger, 94
Arrival Time Property, 127 Draw Queue, 84
Data Table, 124, 134 Drawing
Entities Per Arrival, 127 Create Symbol. See Create Symbol
Repeat Arrival Pattern, 127 Dynamic Selection Rule
Arrival Mode. See Source Arrival Mode Server, 130
AssociatedStation Functions, 229
Capacity, 229
AssociatedStationload, 116 E
AssociatedStationLoad, 229 Elements, 91
AssociatedStationOverload, 229 Failure
Definition, 305
B Material, 143, 144
Monitor, 344, 396
BasicNode, 35, 56 Output Statistic, 132, 144, 274
Converting to TransferNode, 192 Process, 328
Definition, 56, 60 RoutingGroup, 312
External View – Input Node, 292, 369 State Statistic, 144, 146, 156
Batch Members Queue. See Model Entity Batch Members Station, 305, 366, 388
Defintion, 305
Storage, 122, 296
C Tally, 127
Candidate, 83 Tally Statistic, 97, 98, 156, 384
.Entity.Priority, 236, 303 Timer, 147, 156, 281
.Is, 253 Definition, 305
.ModelEntity.ID, 279 Emptier, 359

416 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Ending Time. See Simulation: Ending Time Flow Source, 347, 348
Event, 353 FlowContainer
Creating New, 354 Contents.Volume, 349
Fire, 356 Contents.Volume.Rate, 349
Monitor, 149, 345 FlowToItemConverter, 359
Station Events Function, 89, 276
Entered, 323, 389
Exited, 330
Timer, 147, 159, 373
G
TransferNode Events, 312 Google 3D Warehouse, 210
Wait for, 355
Zero-Time, 239, 308
Experiment, 36, 274 H
Input Analysis, 40, 181 Half-width
Tornado Chart, 40 Estimating Number of Replications, 36
Tornado Chart, 183
KN Select Best Scenario. See KN Select the Best Scenario
Pivot Grid, 38, 395 I
Ranking and Selection (KN), 164 Input Analysis. See Experiment: Input Analysis
Raw Data, 39 Input Sensitivity. See Experiment: Input Analysis
Reports, 38 Is, 256
Response Chart, 162 Entity, 128
Response Results, 38 ModelEntity, 256
Responses, 38, 162 IsParked, 260
Scenarios, 37, 132, 161 ItemstoFlowConverter, 359
SMORE, 39, 162
Explained, 132
Subset Selection, 163 K
Warm-up Period, 182
KN Select the Best Scenario Algorithm, 164, 167, 288
Expression Editor, 25
External View, 321
InputNode, 293, 369 L
Junction, 388
OutputNode, 293, 369 Link
SubModel, 292 Conveyor. See Conveyor
Visible Button, 367 Path. See Path
Selection Weight, 110
TimePath. See TimePath
F List, 139
Accessing List Enumerations, 390
Failure, 202
Data Table Column, 215
Processing Count Based, 114
Node, 82, 113, 115, 228
Types
Objects
Calendar Time Based, 112
Resources, 207
Event Count Based, 112
Standard Property, 387
Processing Count Based, 112
String, 386
Processing Time Based, 112
Strings, 139
Types Explained, 111
Lookup Table, 88, 275
Filler, 359
Financials, 285
Cost Properties M
Fixed Objects Explained, 285
Model Entities Explained, 286 Material. See Elements Material
Worker and Vehicle Explained, 285 Materials Handling, 187
Used, 286 Math
Flow Connector, 347 Epsilon, 241, 308
Flow Node, 347 If, 128
Initial Maximum Flow Rate, 349 Max, 157, 160, 404
Regulator Initially Enabled, 349 Min, 157
Flow Regulator, 355 Remainder, 52, 282
CurrentMaximumFlowRate, 355 Model, 22
Enabled, 355 Model Properties, 22, 145, 147, 291, 311, 344
Flow Sink, 347, 349 ModelEntity, 22, 23, 56
Batch Members Queue, 84
More Subclassing: Process Planning/Project Management 417

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Initial Desired Speed, 55 Triggering Event Name, 133
Initial Sequence, 60 Process Steps
TimeCreated, 97 Assign, 94, 96, 137, 200, 210, 225, 317, 353, 372, 374
TimeInSystem, 97 Change Path Direction, 385
Models Conveyor DesiredSpeed, 201
Airport, 32 Multiple, 147, 159, 200, 375
Airport Revisited, 42 Consume, 147
Assembly of Circuit Boards, 78 Create, 159, 237, 374, 390, 391
Call Centers, 271 Decide, 98, 212, 217, 371, 374, 380
Cellular Manufacturing, 290 Condition Based, 99, 159, 201
Circuit Board Assembly Reconsidered, 93 Delay, 100, 101, 200, 323, 358
Delay Object, 311 Zero Time Event, 308
DMV, 105 Destroy, 99, 160, 375, 376, 390
Gas Station, 350 End Activity, 414
Ice Cream Store, 23 End Run, 131, 134, 401
Kitting Process, 135 End Transfer, 323, 370, 389
Manufacturing Cell, 55, 187 Execute, 277, 307, 346, 371, 374
More Vet Clinic Operations, 227 Addon Process Trigger, 329
Process Planning/Project Management, 383 Fail, 305
Simple Tank, 343 Fire, 356
Supply Chain Insert, 122, 296
Advanced Model, 364 Interrupt, 358
Simple, 153 Notify, 401
TransferLine, 199 Produce, 147, 150
Warehouse Pickup, 302 Release, 100, 101, 373
Monitor, 148 Reserving /UnreservingCapacity, 258
Crossing State Change. See Elements Monitor To List, 209, 213
Worker, 234
Remove, 122, 279, 296, 357
N Repair, 305
NHPP. See Rate Table Reserve, 257
Node Ride, 254
BasicNode. See BasicNode Route, 312
ParkingQueue, 189 Search, 252, 356, 397
TransferNode. See TransferNode ObjectList, 223
Properties Explained, 222, 252
Queue State, 278
O SeizedObjects, 252
OptQuest, 166, 285 Seize, 100, 101, 370
Constraints, 169, 171, 288 From List, 208, 213
Multi-objective Weighted, 169 Number of Objects, 213
Pattern Frontier, 170 Worker, 233
Responses, 286 Set Node, 380
Thresholds, 169 Set Row, 137, 210, 220
Output Statistic. See Elements:Output Statistics Tally, 97, 99, 314, 371, 386, 396
Transfer, 160, 237, 323, 357, 374, 390
UnBatch, 99
P Unreserve, 257
Panels VisitNode, 314
Navigation Panel, 22 Wait, 354
Standard Library, 23 Processes
Path, 57 Add-On Process Trigger
Drawn to Scale, 33 Combiner
Drawn To Scale, 57 Before Processing, 101
Selection Weight, 51 Monitor
Paths, 43 On Event, 345, 396
Bidirectional, 187 Node
Selection Weight Property, 43 Entered, 201, 318
Pivot Grid. See Experiment-Pivot Grid Exited, 201, 289, 318
Process Path
Tokenized, 120 Entered, 201
Triggering Event Condition, 133 ReachedEnd, 150
Trailing Edge Entered, 109
418 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Server Create New Property, 37
Before Processing, 208 Definition, 50
Entered, 95, 277 Set Referenced Property, 37
Exited, 96 Spreadsheet View, 199
Failed, 202 User-defined, 83
Processed, 157, 209, 217, 234, 291, 304, 355
Processing, 291, 304, 354
Repaired, 202
Q
Sink Queues. See Animated Queues
Destroying Entity, 99, 131
Entered, 353
Source R
CreatedEntity, 137, 210, 352 Random
Exited, 94 Binomial, 138
SupplyInv Discrete Distribution, 47
ReorderExited, 380 Ranking and Selection, 164
WorkStation Rate Table, 45, 273
FinishedGoodOperation, 160 Referenced Property, 83, 215
FinishedSetup, 143 Reliability. See Failure
Create Process, 96, 101 Reneging, 271, 356
Delay Object Repeating Property Editor, 81
OnEntered, 323 Bill of Materials, 144
OnExited, 330 Release, 209
InventoryReview, 159, 373 Seize, 208
Junction Object State Variable Assignment, 85
OnEnteringProject, 389 Resource, 87, 129
On_Off_Entered, 200 Initial Capacity, 87
Release Associate, 303 Ranking Rule, 129
RenegeProcess, 278 Reserving, 257
RespondToTimerEvent, 147, 282 Results, 29
Seize Associate, 303 Categories, 30
SetTimeClock, 317 Capacity, 30
Supplychain Object Content, 30
OnEnteredOrdering, 370 FlowTime, 30
OnExitedOrderingProcess, 373 HoldingTime, 30
Modifying ResourceState, 30
Override, 313, 373, 375, 376 Throughput, 30
OnReachedEnd, 385 Pivot Grid/Table, 30, 36
Restore, 314 Resource utilization, 30
OnEnteredInputBuffer, 305 Run
OnEnteredProcessing, 306 TimeNow, 52, 97, 282
OnRunInitialized, 157, 346 Run Setup, 26
Adding, 157, 346
TimePath
All Defined, 385 S
Tokenized, 102, 150 Schedules. See Work Schedules
Tokenized, 234 Secondary Resources, 87, 243
TransferNode SeizedResources
All Defined, 312 All Functions, 217
Triggering Event, 147 LastItem, 242
Trigging Event, 159 NumberItems, 218
Project Management Used in Search Step, 253
Critical Path, 383 Separator, 92
Earliest Finish, 383 Server, 23, 56
Latest Finish, 383 Add-On Process Trigger. See Process Add-On
Percent CP(Critical Path), 383 Capacity
Slack, 383 Workschedules. See WorkSchedules
Properties Capacity Type, 58
Element Reference, 328 Dynamic Selection Rule, 130
Object Reference, 215 InputBuffer, 237
Repeat Group, 332 InputBuffer.Contents, 26
Standard Property, 48 Process Types, 261
Property
More Subclassing: Process Planning/Project Management 419

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Processing Time, 89 On Entering State Assignments, 91
Processing.Contents, 26 String, 117
Reliability Logic, 202 Vector, 117
Resource States Explained, 105 State Variables
States Explained, 105 Vector, 70
Static Ranking Rule, 224 Station
Station EntryQueue, 321, 335
InputBuffer, 105 Statistics
OutputBuffer, 105 Observation-Based, 90
Processing, 105 Output. See Elements Output Statistic
Stations Explained, 105, 305 Results/Reports. See Results
Subclassing. See Subclassing State, 68
Taken Apart and Features Explained, 305 State Statistic, 90, See Elements State Statistic
Task Sequence. See Task Sequence Tally, 68, 90, 108, See Elements Tally
Zero-Time Processing Pickup Point, 196 Time-persistent, 90
Setup Times Status Labels. See Animation
Sequence Dependent. See WorkStation:Setup Storage, 122
SIMIO String
Start page, 21 Format, 161, 317
Simulation Subclassing, 305
EndingTime, 45 Delay Object, 311
Input Analysis. See Experiment:Input Analysis Derived/Specialized, 311
Output Analysis, 173 Server, 305
Replication Number, 176 TimePath, 384
Starting Time, 45 TransferNode, 311, 393
Steady State, 173 Workstation, 365
Terminating, 173 Symbol
Warmup period, 178 Create New, 79
Sink, 23
Add-On Process Trigger. See Processes Add-On
Statistic Collection, 92, 277
T
SMORE, 133, See Experiment SMORE Tables
Source, 23, 56, 126 Data Table, 48, 62, 209, 219, 281, See
Arrival Mode Appointment Table, 124, 134
Arrival Table. See Appointment Schedules Arrival Table Absolute, 124
Interarrival, 25, 33 Arrival Table Relative, 134
On Event. See Source Arrival Mode Automatic Creation of Element Columns, 74, 76
Time Varying Arrival Rate, 46, 273 Initial Property Values, 74
Before Exiting State Assignments, 47 Automatic Row, 63
Before Exiting State Assignments, 85 Random Row, 64
Entities Per Arrival, 57 Related Table
Initial Capacity, 37 Foreighn Key, 65
Maximum Time, 131 Set Column as Key, 65
Single Source vs. Multiple Source, 61 Related Tables
Source Object, 25 Foreign Key, 142
State Assignments, 94, 155 Related Tables, 64, 141
Stopping Conditions, 131 Related Tables, 219
Table Reference Assignment, 63 Resource List Column, 215
Time Varying Arrival Rate, 46 Row Specification, 49
Standard Property, 48 Search. See Process Steps Search, See Process Steps
Integer, 48 Search
State, 331 Tally Statistic Element Reference, 71, 72
String, 209 Data Tables
Starting Time. See Simulation:Starting Time Child Table, 141
State Assignments. See State Variable Sequence Table, 59, 62
Creating your own, 331 By Sequence TransferNode, 60
State Statistics. See Statistics:State Related Table, 64
State Variable, 47, 117 Table Reference Assignment, 63
Before Existing State Assignments, 81, 155, 377 Time Indexed, 283
Definition, 50 TimeIndexedRowFunction, 283
Discrete State, 80 TimeIndexedValueFunction, 283
Integer, 136 Tally Statistics. See Statistics:Tally
String, 136
420 Simulation Modeling with SIMIO: A Workbook

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]
 Tank, 347, 348 Selection Goal, 116
Tank Empty Add-on Process Trigger, 350 Setting Traveler Capacity, 119
Task Sequence, 261 Subclassing, 311
Branch Types, 264 TransferNode Properties, 44
Precedence Rules, 263 Trimble 3D Warehouse, 33
Process Type
Process Name, 266
Specific Time, 266
V
SubModel, 266 Vehicle, 188, 232
TimePath, 24, 82, 316 Fixed Route, 197
Random Time, 135, 378 Free Space Travel, 197
Reached End Process Trigger, 150 Idle Action, 189
Travel Time, 83 Initial Desired Speed, 189
Timer Element. See Elements Timer Initial Node(Home), 189
Token On Demand, 197
Add New, 120 ParkingStation.Contents, 190, 191
Adding New, 102 Ride Capacity, 189
Associated Entity, 251 RIDE STATION queue, 190
Multiple Tokens, 277 Task Selection Strategy, 257
Creating New Ones, 289 Vehicles, 187
Custom, 102 View, 26
Explanation, 93 2-D, 26
Return Value, 223, 252 3-D, 26
Returned from Search, 222, 252
tornado chart. See Experiment:Input Analysis:Tornado Chart
Transfernode W
Reserving the transporter for later use, 257 Work Schedules, 51, 58, 273, 299
Routing DAY PATTERNS, 51
Blocked Destination Rule, 231 Start Date Explanation, 77
Choosing From List, 231 Using Properties, 284
Selection Condition, 231 Worker, 232
Selection Expression, 231 Dynamic Selection Rule, 236
Selection Goal, 231 Transporter, 245
TransferNode, 35, 44, 187 Workstation, 135, 154
All Processes, 312 Changeover Matrix, 139
BySequence, 60 Operation Quantity, 139, 156
Decision Point, 52 Other Requirements, 145
Definition, 56, 60 Processing Batch Size, 139, 156
Entered Add-On Process, 201 Setup Time
Entity Destination Type, 60, 82 Fixed, 136
Entity Destination Type- Select from List, 83 Setup Times
Entity Destination Type-Shortest Path, 115 Sequence Dependent, 138
Events, 313 Random, 141
Exited Add-On Process, 201 Subclassing, 364
External View – Output Node, 293, 369 Teardown, 136
Inherited Properties, 312
Outbound Link Rule, 51
ParkingStation, 312 Z
Ride On Transporter, 188, 193, 247 Zero-Time Events. See Event-Zero Time
Ride Process Step, 254
Selection Expression, 83

More Subclassing: Process Planning/Project Management 421

Licensed to: Patricio Matus Carrillo - Avenida Los Carrera 2114 Concepción 4030000 - [email protected]