Factory Physics

Methods for Efficient Project Management in a Scientific

Environment—Part I: Operations Science
Mark L. Spearman, PhD
26 October 2021

Copyright © 2021 Project Production Institute. All Rights Reserved.

What is Factory Physics?

Copyright © 2021 Project Production Institute. All Rights Reserved.

1
 Description of the way factories behave
Title of a book first published in 1996
Name of a company that offers services to manufacturing companies

Copyright © 2021 Project Production Institute. All Rights Reserved.

Why Factory Physics?

Copyright © 2020 Project Production Institute. All Rights Reserved.

2
 Context

Copyright © 2021 Project Production Institute. All Rights Reserved.

Supply Flow

E F D I

E F D I

E F D I

E F D I

Copyright © 2021 Project Production Institute. All Rights Reserved.

3
 Definitions

Copyright © 2021 Project Production Institute. All Rights Reserved.

𝑟 = Capacity: max units/tasks that can be done per unit time

TH = Throughput: units/tasks completed per unit time
u = Utilization: fraction of capacity that is used, 𝑢 = TH/𝑟
CT = Cycle Time: time from start to finish of a unit/task
WIP = the number of incomplete units/tasks

Copyright © 2021 Project Production Institute. All Rights Reserved.

4
 Inventory, WIP, Stock and Queues

Copyright © 2021 Project Production Institute. All Rights Reserved.

Copyright © 2021 Project Production Institute. All Rights Reserved.

5
 Modeling
Analytic and Queueing
Monte Carlo Discrete Event Simulation

Copyright © 2021 Project Production Institute. All Rights Reserved.

Copyright © 2021 Project Production Institute. All Rights Reserved.

6
 Fundamentals

Copyright © 2021 Project Production Institute. All Rights Reserved.

Little’s Law:
WIP=CT×TH
TH is demand
WIP is “visible cycle time”

Copyright © 2021 Project Production Institute. All Rights Reserved.

7
 What is “Pull?”
WIP and TH are ‘leading’ indicators, CT lags
Pull controls WIP, Push controls TH
Pull is more robust than Push

Copyright © 2021 Project Production Institute. All Rights Reserved.

Performance Curves

Copyright © 2021 Project Production Institute. All Rights Reserved.

8
 FLOW STOCK

Copyright © 2021 Project Production Institute. All Rights Reserved.

Variability
Demand and Production are never synchronized
There is always variability!

Copyright © 2021 Project Production Institute. All Rights Reserved.

9
 Variability in a production system will be buffered with some combination of:
1. Inventory
2. Time
3. Capacity

Copyright © 2021 Project Production Institute. All Rights Reserved.

Operations Science

Copyright © 2021 Project Production Institute. All Rights Reserved.

10
 Almost everything we do involves operations
Manufacturing
Transportation
Medical services
Project execution

Copyright © 2021 Project Production Institute. All Rights Reserved.

Operations Science

Science, not just mathematical models

Copyright © 2021 Project Production Institute. All Rights Reserved.

11
 Little’s Law = Little’s Tautology
Does throughput always go up with additional WIP?
How do the three buffers interact?

Copyright © 2021 Project Production Institute. All Rights Reserved.

Operations Research and Operation Science

Adan, I., J. Van der Wal. 1989. Monotonicity of the Throughput of a Closed Spearman, M.L., W.J. Hopp. 2021 The Case for a Unified Science of Operations,
Queueing Network in the Number of Jobs. Operations Research, 5 (6), 953- Production and Operations Management, 30 (3), 802-814
957.

12
 Operations Science

The study of the transformation of

entities using resources to create
and distribute goods and services
that satisfy a given demand

Copyright © 2021 Project Production Institute. All Rights Reserved.

Operations Science describes

Demand processes
Resource utilization
Variability due to randomness and lack of information
Buffering and synchronization
Work in process, cycle time, throughput
Stocks, backorders and lost demand

Copyright © 2021 Project Production Institute. All Rights Reserved.

13
Process

Example: Engineer, Fabricate, Deliver, Install

Process begins by accumulating all needed resources
Proceeds through a sequence of operations

Demand

Represents the desire of a “customer” for an entity or a set of entities with

particular attributes
The “customer” can be individuals or the downstream process or the sponsor of
a project
Entities demanded can be physical, information, services
Demand is always satisfied by a process

14
 Projects
Creation of a unique device, structure, software, etc.
The creation is unique but composed of many repetitive processes

Copyright © 2021 Project Production Institute. All Rights Reserved.

15
 Maintain the Rates  Make the Dates

Schedule = Demand / Should Production System = Supply / Will

TAKT

TH

Dates & Progress Rates / Throughput

Copyright © 2020 Project Production Institute. All Rights Reserved.

Variability and Buffer Interactions

Copyright © 2021 Project Production Institute. All Rights Reserved.

16
 Simple production-inventory system

𝝁, 𝒄𝟐 𝜆, 𝜓
S

Queue Operation Stock Demand

Queue + Operation + Stock = Basestock = S

(stock can be negative—backorders)

Copyright © 2021 Project Production Institute. All Rights Reserved.

Buffers
Time—demand waits on unit or task
Inventory—units or tasks wait on demand
Capacity—productive ability that exceeds average demand

Copyright © 2021 Project Production Institute. All Rights Reserved.

17
 Time and Inventory Buffers mirror each other
Time-Inventory Buffer

Capacity Buffer mitigates Time-Inventory Buffer

Copyright © 2021 Project Production Institute. All Rights Reserved.

Explore buffer relationships with simulation

Copyright © 2021 Project Production Institute. All Rights Reserved.

18
Copyright © 2021 Project Production Institute. All Rights Reserved.

Time-Inventory Buffer
The standard deviation of lead time demand

Capacity Buffer
Capacity less Average Demand

Copyright © 2021 Project Production Institute. All Rights Reserved.

19
 Let 𝜓 = 𝑐 = 𝑉
Vary V and 𝜇 − 𝜆 while keeping 𝜎 constant

Copyright © 2021 Project Production Institute. All Rights Reserved.

𝜆 = 1, 𝜇 = 1.25 → 1.05
𝜓 = 𝑐 = 0.93 → 0.22

𝜎 ≈ 4.4
Copyright © 2020 Project Production Institute. All Rights Reserved.

20
 𝜎 +𝜎
𝜇−𝜆 𝜎 ≈
2

Copyright © 2021 Project Production Institute. All Rights Reserved.

Inventory /
Time Buffer
Inventory /
Time Buffer

Capacity Buffer
Capacity Buffer

Copyright © 2021 Project Production Institute. All Rights Reserved.

21
 Can you have an inventory of completed projects?

Copyright © 2021 Project Production Institute. All Rights Reserved.

Cycle Time

Efficiency (utilization)

Copyright © 2021 Project Production Institute. All Rights Reserved.

22
Copyright © 2021 Project Production Institute. All Rights Reserved.

Copyright © 2021 Project Production Institute. All Rights Reserved.

23
 Must have a capacity buffer to avoid the “time buffer”

Being overly “efficient” means being late

Copyright © 2021 Project Production Institute. All Rights Reserved.

Project Duration

Copyright © 2021 Project Production Institute. All Rights Reserved.

24
 Too much WIP
long cycle times
inefficient production
Too little WIP
insufficient production

Copyright © 2020 Project Production Institute. All Rights Reserved.

Optimal Typical
WIP Level WIP Level

Copyright © 2021 Project Production Institute. All Rights Reserved.

25
 Routings
Rates
Yields Production System Model
Downtime
Etc.
Production System Model
Little’s Law:

Cycle Time
CT = WIP / TH
Utilization

Throughput
Cycle Time Formula:

CT = RPT + BT + MT + QT + SDT + WTMT + PTB

WIP

VUT Equation:
RPT = PT + ST + DT

Cycle Time
BT = (Waiting for Batch) + (Waiting in Batch)

Copyright Strategic Project Solutions, Inc.

WIP

Data Science Operations Science

© B.O. Tayo

Stock
Work-in-Process
Parameters Batch
Capacity
Cashflow

Production System Control Policies

Copyright © 2021 Project Production Institute. All Rights Reserved.

Questions?

Copyright © 2021 Project Production Institute. All Rights Reserved.

26