See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/221525852

Generic simulation of automotive assembly for interoperability testing

Conference Paper in Proceedings - Winter Simulation Conference · December 2007

DOI: 10.1109/WSC.2007.4419702 · Source: DBLP

CITATIONS READS
34 1,384

2 authors, including:

Deogratias Kibira
National Institute of Standards and Technology
54 PUBLICATIONS 653 CITATIONS

SEE PROFILE

All content following this page was uploaded by Deogratias Kibira on 18 July 2015.

The user has requested enhancement of the downloaded file.

Proceedings of the 2007 Winter Simulation Conference
S. G. Henderson, B. Biller, M.-H. Hsieh, J. Shortle, J. D. Tew, and R. R. Barton, eds.

GENERIC SIMULATION OF AUTOMOTIVE ASSEMBLY FOR INTEROPERABILITY TESTING

Deogratias Kibira
Charles R. McLean

Manufacturing Systems Integration Division,

National Institute of Standards and Technology
100 Bureau Drive, MS 8260
Gaithersburg, M.D. 20899, U.S.A.

ABSTRACT impossible for researchers in institutions and universities to

duplicate real life manufacturing systems due to the high
Computer simulation is effective in improving the effi- costs of manufacturing hardware and software. Addition-
ciency of manufacturing system design, operation, and ally, they do not have unlimited access to real
maintenance. Most simulation models are usually tailored manufacturing systems since this would interfere with pro-
to address a narrow set of industrial issues, e.g., the intro- duction activities. Companies are also reluctant to supply
duction of a new product. If generic data-driven much detail about plant operations since some of the in-
simulations could be developed they would be reusable for formation is confidential.
wider application including interoperability testing of stan- To circumvent this difficulty, researchers are con-
dards for exchange of data across the supply chain in structing generic, data-driven simulation models of
manufacturing. To facilitate future interoperability testing manufacturing systems to facilitate current and future,
and training, scientists at the National Institute of Stan- training, experimentation, and testing of interoperability of
dards and Technology are currently developing distributed, software. By generic simulation the model can be recon-
integrated manufacturing simulations for automotive figured for many situations in automotive assembly.
manufacturing. These simulations are being developed at Generic simulation identifies common model input and
four different levels: the supply chain, the assembly plant, output data interfaces that could be standardized for par-
the engineering systems, and the shop floor level. This pa- ticular modeling level and simulation case studies
per describes the development of a simulation model of the (McLean and Shao, 2003). Case study templates can be
final assembly plant. Future efforts will increase the versa- developed that are generic for a specific domain such as
tility of the model, run it on neutral data and extend scheduling, plant layout, materials handling, new equip-
integration with supply chain simulation. ment. This paper is not based on any specific automotive
final assembly plant but is a step in that direction. By being
1 INTRODUCTION data-driven, it will have information relating to part-
subcomponent association, process definition, part routing,
Automotive manufacturing is complex and includes the and initial inventory levels defined and specified outside
coordination of design and manufacturing activities be- the model. Such information would then be read into the
tween many companies. The process involves a number of simulation at initialization time during a run. This enables
operations, which require assembling together thousands of modification of production operations of the final assembly
fabricated, and purchased components, subassemblies, and plant with minimal changes to the simulation model.
systems. Purchased components are outsourced from Previously, the general application of simulation in
many supplying companies who generally use different automotive manufacturing centers on investigation of op-
data formats, which are not always compatible. Yet this erational options of different shops used in the process.
product data must be shared among many companies in- For example Lohrer (1997) identifies that simulation is
volved in the production activities. This lack of software usually applied to investigate body, paint, and
interoperability among different companies along the sup- trim/chassis/final assembly shops. There is also literature
ply chain causes major cost increases in the manufacturing on engine block casting, machining, and assembly. The
industry. Researchers would need to test and evaluate the focus of investigation in these applications is often to in-
suitability and effectiveness of existing and candidate stan- vestigate the effect on new tooling, materials delivery
dards for application to specific manufacturing areas. It is systems, the impact of failures, etc. A dynamic operating
 Kibira and McLean

algorithm for the painted body storage in an automotive grammable logic controller may be tested with an
manufacturing plant is presented in Moon et al. (2005). It emulated conveyor system before the physical
investigates grouping cars of the same color together to re- conveyor system is installed or even delivered.
duce changeover costs. Simulation has also been used to • Evaluate the capability of the delivered process,
determine the cycle time of the robots and buffer sizes be- system ,or design to meet interface specifications.
tween sub-lines of body shop Moon et al. (2006). Some • Perform conformance and acceptance testing us-
applications also focus on the supply chain such as Jain et ing simulations to create the specified range of
al. (2005). inputs for a delivered system or process.
Currently, there are no dynamic, manufacturing ori- • Evaluate whether new systems, processes, or de-
ented testing facilities to evaluate the suitability of signs meet performance requirements and
standards for selected applications. There also lacks ways specifications. For example, test programs for
to identify and resolve conflicts between standards, and robots and materials handling systems using simu-
evaluate compliance of vendor implementations with stan- lations.
dards. Dynamic testing capability would enable the live • Develop metrics to allow the comparison of pre-
testing of multiple independently operating manufacturing dicted performance against “best in class”
subsystems. The linkage between subsystems would be benchmarks to support continuous improvement
various interface standards and protocols developed by dif- of manufacturing operations
ferent standards organizations. As such, virtual
manufacturing environments with data-driven simulation 2 MOTOR VEHICLE ASSEMBLY PROCESS
could be used by manufacturing companies for training,
experimentation, and testing purposes. The simulation of This section describes the manufacturing process and the
the assembly plant will include such issues as facility lay- development of the assembly simulation model. Informa-
outs, materials handling, and system schematics for major tion about motor vehicle manufacturing was obtained from
production shops. It will also be concerned with the ex- published literature and reports of visits to the Volvo motor
change of data such as bills of materials, configuration, lot vehicle plant in Gothenburg (Sweden) and the General Mo-
sizing. tors plant for Cadillac/Buick vehicles in Detroit, Michigan.
While the overall goal of the Virtual Manufacturing However, the simulation model does not represent the pro-
Environment (VME) project is to provide interoperability duction system of either plant. A typical automotive
testing support to software developers, manufacturers, re- assembly plant has more than one thousand stations. But
searchers, and standards organizations using a virtual because of the desire to simulate the entire plant rather than
reality simulation environment, the objectives of the work part of the process, it became necessary to consolidate
presented here are: processes to reduce this number.
• Identify facilities, systems, operations, parts, and The automotive production process consists of three
processes in automotive manufacturing assembly major sections: the body shop, the paint shop, and the trim
to develop the model. assembly shop. Other sections are the power train assem-
• Develop the simulation model of the final assem- bly (consisting of the engine, gearbox, clutch, and
bly plant. transmission), and the press shop if body parts are stamped
• Integrate the assembly simulation model with at the plant. There is also a final testing process where ve-
other simulations using High Level Architecture. hicles are checked for water tightness and a stationary road
• Carry out interoperability testing using test case test.
data.
Some examples of possible simulation-based testing 2.1 The Body Shop
applications include:
• Evaluate effectiveness of new interface standards The first stage in the production of a motor vehicle is the
and protocols to meet manufacturing industry fabrication and assembly of what is called the “white
needs. body” or “body in white” of the car. The major compo-
• Evaluate conflicts and inconsistencies between nents of the automotive body are the underbody (or
standards developed by different organizations. sometimes called the floor pan), body sides, framing, hood,
• Perform interoperability testing with models of trunk lid, doors, and roof. These are produced by separate
systems being integrated. For example, a model robotic cells. The underbody is in turn made up of the
of a robot controller may be integrated with a front, middle, and rear sections. The front section is made
model of the robot for testing purposes to ensure up of the engine compartment and mounting for instrument
interoperability. panel. The middle section is the under floor of the passen-
• Perform interoperability testing with emulated ger compartment while the rear section comprises the
physical equipment. For example, a physical pro- trunk. The sections are produced from stamped parts. The
 Kibira and McLean

underbodies are usually bar-coded at this stage to indicate Most of the above processes are done by robots except
body type. Then they are transported to another section where human intervention is required. In general, air flow
where the sides are attached to the underbody. There is and water reservoirs are used to carry away excess paint.
usually a storage space for underbodies at the start of the Small defects in the paint finishes are generally corrected
following section. manually. After painting the body is sent to the final trim
The framing and body structures are further welded to assembly shop. One of the typical problems encountered is
the underbody, and after which the roof attached by weld- the sorting of the incoming vehicles to minimize color
ing. The car body then begins taking shape. The body changes. Although robots can change colors very quickly,
sides consisting of the entire side from the trunk to the flushing the entire shop to change from painting one color
hood, except the doors are assembled at separate stations to another takes a setup time. Therefore, it is desirable to
and transported to the body side assembly area. Cross roof sequence same color vehicles one after the other – a con-
supports are also welded on the side panels. The roofs are cept called color blocking. This necessitates a temporary
sometimes bought or stamped and assembled at the shop. storage for white bodies before the paint process. While
Some roofs are the “open roof” type, some are closed changing colors, it is usual that white or lighter colors pre-
roofs. At another station the doors, hood, and trunk lid are cede darker colors rather that the other way round.
also assembled to the body. Another point, according to Ulgen et al. (1998), is the per-
centage of painted cars passing the quality control check. It
2.2 The Paint Shop is called the yield of the painting process. Yield can be as
low as 65% and the chances of rework are high. This can
This is the shop where the body in white is painted and be a major source of process variation. The process ahead
given the final color and texture required in the final sale- of the paint process should have sufficient storage to pre-
able vehicle. There are often a number of parallel paint vent blockage.
lines. Typically many processes are involved here. The
stages in the paint sequence are invariably as follows: 2.3 The Trim Assembly Shop
• Degrease – clean any grease on vehicle bodies
that stuck as a result of the body forming process. Trim assembly is where all parts and assemblies needed for
• Phosphate wash – wash any oil on the body so a vehicle to move as well as other conveniences are assem-
that the paint can stick to the body. bled into the body. There is usually a separate shop for
• Dry – bake the body in an oven to dry. fabrication and assembly of the chassis and power system
• Electro paint – dip the body into a tank containing of the vehicle. The engine is first fit with various features
the paint and apply an electric charge so that the according to required specifications (engine dressing) and
paint can stick to the body. assembled together with the transmission system (clutch,
• Clean – clean the body of any dirt. gear box, propeller shaft, etc). The power system is at-
• Dry – bake the body in an oven to dry. tached to the chassis onto which axles, suspension,
• Pre-seal – seal off or plug any hole left in the exhaust, steering, and brake system have already been as-
body and perform some touch up grinding work. sembled. Some parts are usually assembled into the body
before it is merged with the power system; the process is
• Undercoat paint – apply a second coat of paint
called “body drop” since it is the body that is usually low-
which determines the final color of the vehicle.
ered onto the chassis/power system. In many cases the
• Dry – automobile body is again in an oven.
doors are removed at the beginning of the trim assembly.
• Light application of sand – remove any dust and After various parts are assembled at a separate station into
manual cleaning or using a feather duster ma- the doors, they are re-attached to the car at a later stage.
chine. Removing doors before trim assembly allows easy access
• Primer manual paint – manually paint the inside to the inside of the vehicle and to reduce possible damage
of the vehicle where robots may not easily reach. to doors. In some plants the doors are not removed from
• Outside coat painting using robots – apply the un- the body during trim assembly. In this case they have to be
dercoat paint. left open through most of trim assembly, requiring larger
• Dry – bake the body once again dried in an oven. assembly space.
• Quality control check – ensure work is done well Many older vehicles had separate underlying stiffening
so far. The bodies that do not satisfy required structures and bodies, the body housing the passengers.
specifications are re-done. This design requires more materials and results in a heavier
• Top coat paint application – apply the final color car, thus raising costs. Today most cars are manufactured
paint. with a unibody spaceframe chassis. This means that the
• Dry – final baking of the body in an oven to dry. body itself is constructed such that it provides the stiffness
required by the vehicle. In this case the body is dropped to
 Kibira and McLean

the axles onto which the power, suspension, steering, inserted The underside of the body is also assembled at
brake, and fuel delivery systems are already assembled. In trim assembly. The fuel delivery, exhaust systems, etc. are
some cases the hood is removed before the power system is firmly clamped to the body. There is also where various
merged with the body. This allows easier access to the en- parts not directly attached to the engine are finally inserted.
gine compartment and reduces possible damage. Most The tires are finally bolted onto the vehicle.
trim assembly operations are carried out manually. Typi- Door assemblies with windows, switches, and systems
cally, there is a worker or two on either side of the line at a for closing and opening windows, electrical work, side
station. There is sufficient space on which to work, mirrors, arm rest, etc. and hood are re-attached in the last
equipment and tools specific to the station, and there are stages. Fluids (brake fluid, engine oil, transmission system
racks or bins on which the parts are stored for assembly. fluid, water, gas) are added to the reservoirs. The vehicle
This is where a variety of configuration options can be is then tested for water tightness. The final stationary road
made by the customer. test, carried out inside the plant, ensures the vehicle and
It is in the trim assembly shop that electrical wiring is metering systems are working correctly.
added into the body and engine compartments, weather Figure 1 summarizes the description of the automotive
proofing, carpets and floor mats, dashboard and instrument fabrication and assembly process. This formed the basis for
panel, steering wheel, gear lever, handbrake, and pedals, the simulation model development of the next section.
vinyl top, bumpers, inside lighting, outside lighting, indica-
tor lamps, mirrors, windshield, wipers, seats, etc. are

Figure 1. Basic processes in an automotive final assembly plant

 Kibira and McLean

production level is 6,000 cars. Once a car is fully assem-

3 AUTOMOTIVE ASSEMBLY SIMULATION
bled and has exited through the last stage on the line it is
compared with the first order in the sequence. The total
3.1 Product Options
number of cars required to complete that order is decre-
mented by one if the current order of car options is the
There is one model of car to be manufactured but can
same as those of the exiting car. If not, the next order in
have many options. The options on the car are that it can
the sequence is considered. When the total number of
be either two door or four door. A two-door car will re-
units in an order are produced a message is sent to the
quire a different body side panel from a four-door car.
supply chain simulation. In the current version of the
Additionally, the car can be either open roof or ordinary
simulation model, an order comprises only one type of
roof (closed). Either a two or four door car can have open
product.
or ordinary roof. Thus, there are four options at the “body
in white” car stage. There are three color options avail-
3.3 Delivery of Parts
able at the paint shop. This totals to twelve options as
shown in Figure 2.
When production data is received the model creates the
parts required using a bill-of-materials for the required
Type of Type of Color configuration of car. The parts which differ for either
car roof Blue Grey Tan two-doors or four-doors or open roof and closed roof op-
tions are the underbody parts, side panels, doors, roofs,
Closed and trunk lids. When there are no orders to process the
roof 1 2 3 model does not have parts. The components created are
4- door car
Open joined into sub assemblies and transported by the materi-
roof 4 5 6 als handling system to stations where they are assembled
Closed to the main body of the automobile.
roof 7 8 9
2- door car 3.4 The Simulation Model
Open
roof 10 11 12
The simulation model of the assembly line operations was
Figure 2. Description of options modeled in the assembly developed in Delmia QUEST. QUEST is a discrete event
line simulation simulation tool with three dimensional visualization, im-
port, and export of data capabilities. A section of the
3.2 Production Data simulation model is shown in Figure 3. In addition to ex-
ternal interface mechanisms i.e. files and user-defined
The simulation model receives data about required pro- popups, QUEST uses ‘sockets’ as a bridge to communi-
duction in terms of type of car and quantities from the cate and exchange data with other simulations systems.
supply chain simulation. This information about what is The workstations are arranged and connected with an
to be made and in what quantities is not known before the appropriate materials handling system using features
start of a simulation run but rather this information is re- available in the simulation tool. Workstations are mod-
ceived dynamically during the course of a simulation run eled as QUEST machines. In the body shop the conveyor
as would be the case in a real world operation. The orders is used as the materials handling system to move materi-
are sequenced for processing on a first-come first-serve als between stations. All components were originally
basis. In future enhancements more sophisticated se- modeled as separate entities that are joined to form a new
quencing procedures will be developed at the beginning part which is conveyed to the next workstation. In the
of the body, the paint, and the trim assembly shops. paint shop the power and free system and conveyors are
Variability of process time is included in the model. used for materials handling. The stations are modeled as
A probability distribution is also used to model failures at either machines or conveyor decision points. In the trim
test stations. In automotive manufacturing production shop, conveyors and power and free conveying systems
rates are usually described in terms of cars per hour. The are likewise used. Simulation control language (SCL),
production rate of the assembly plant is set at approxi- encoded in the process logic of workstations, is used to
mately 60 cars/hour. The production line is assumed to select the appropriate process to execute when a part ar-
be balanced such that the sum of cycle time and transfer rives. Different displays, based on appearance of motor
time between stations is approximately one minute. Con- vehicles during subassembly, are used for each subassem-
sidering a 10 hour shift and 2 shifts per day the bly and component. Probability distributions of failure are
production rate of the plant should be set at 1200 cars per used at test quality control check stations after the body
day. Considering 5 working days a week, the weekly and paint shops.
 Kibira and McLean

The CAD feature of QUEST is used to model parts <?xml version="1.0" encoding="utf-8" ?>
and workstations in the simulation model. Others which - <AcknowledgeShipment xmlns =
have complex shapes, such as the car body, are obtained "http://www.openapplications.org/oagis/9"
by importing bought models of cars. The number of units xmlns:xsi=http://www.w3.org/2001/XMLSchema-instance
xsi:schemaLocation="http://www.openapplications.org/oagis/9
contained in buffers at stations is displayed. The number ../../BODs/Developer/AcknowledgeShipment.xsd">
of stations has been reduced from 1,200 for a typical - <ApplicationArea>
automotive assembly plant to about 60 in the simulation - <Sender>
model. For painting, it is assumed that there are three <LogicalID>123</LogicalID>
parallel painting lines all carrying out the same set of op- </Sender>
erations and the materials handling system is made up of <CreationDateTime>2006-08-13</CreationDateTime>
conveyors and power and free systems. Cars are trans- <BODID>CurOrder->orderId</BODID>
ported between stations in the paint and the final trim </ApplicationArea>
assembly shops using power-and-free and ordinary con- - <DataArea>
- <Shipment>
veyor systems. The automotive final simulation model - <ShipmentHeader>
has been integrated with the supply chain using the High <ActualDeliveryDateTime>2006-08-13
Level Architecture (HLA) as the Run Time Infrastructure </ActualDeliveryDateTime>
(RTI) (McLean et. al., 2005). A demonstration of this in- - <ShipFromParty category="Organization">
tegration has been carried out. - <PartyIDs>
The simulation includes associations and interactions <ID>123</ID>
between the supply chain nodes. Production order re- </PartyIDs>
quirements and other messages encoded in eXtensible <Name>MyCarFactory</Name>
- <Location>
Markup Language (XML) (Goldfarb, 2000) are passed
<ID>123</ID>
from the supply chain simulation to the assembly simula- </Location>
tion. These are the messages exchanged via sockets since </ShipFromParty>
QUEST does not communicate directly with other sys- - <ShipToParty category="Organization">
tems. Order completion and shipping messages from the - <PartyIDs>
assembly simulation to the supply chain components are <ID>123</ID>
likewise transmitted. The interaction messages use data </PartyIDs>
fields consistent with those defined in Open Applications <Name>MyDealer</Name>
Group’s Integration Specification /Automotive Industry - <Location>
<ID>123</ID>
Action Group (OAGIS/AIAG) Business Object Docu-
</Location> CurOrder->shippingInfo
ments (BODs) for Inventory Visibility and </ShipToParty>
Interoperability (IV&I). For example, shipment notifica- </ShipmentHeader>
tions that are sent from the assembly plant to dealers use - <ShipmentItem>
XML messages that are formed using the SyncShip- - <ItemID>
mentSchedule BOD specification (OAGi, 2007). <ID>CurOrder->CarType</ID>
The QUEST simulation reads and sends XML mes- </ItemID>
sages as a single continuous string of characters. The <Description>MyFirstCar</Description>
string has to be searched for the appropriate information it - <PurchaseOrderReference>
- <DocumentID>
contains. For example, car order data on specifications
<ID>123</ID>
and quantities are extracted from the XML message and </DocumentID>
stored in a dynamic SCL list structure. This list is up- </PurchaseOrderReference>
dated with the arrival of a new order when a new one - <ItemSubLine>
arrives. When a particular order is completed it is deleted <Quantity>CurOrder->Quantity</Quantity>
from the list. An example XML shipment message is in- </ItemSubLine>
dicated in the exhibit in Figure 4. Since orders are </ShipmentItem>
received and executed dynamically we will refer to a cur- </Shipment>
rent order (CurOrder). The CurOrder->orderId refers to </DataArea>
the current order identification. MyFactory and </AcknowledgeShipment>
MyDealer refer to the plant and dealer identification. The
CurOrder->shippingInfo in the shipping information. Figure 4: XML Message notifying headquarters of the
Other information are the CurOrder->CarType and completion of a production order.
CurOrder->Quantity, which refer to the type of car and
the quantity that has been shipped.
 Kibira and McLean

4 THE WAY FORWARD Simulation Data (CMSD) specification (CMSD

Product Development Group, 2006) currently under
The project has gone through the first phase of developing development. Such data can be applicable to other
the simulations and integrating with the HLA so that the simulation systems and can also be changed with
simulations can exchange messages. Future plans in the minimal or no change to the simulation model.
development of the generic simulation will increase the • The animation of operations of some workstations
sophistication of the model to make it more reflective of using Delmia IGRIP and incorporate them into the
real production systems, and to handle a wider range of a QUEST model.
exchanged messages. The following are planned to be in-
corporated: 5 CONCLUSION
• The number of workstations will be increased.
• A system using a set of rules to be written in This paper has described the development of an automo-
QUEST simulation control language (SCL) will be tive final assembly simulation model to enable
incorporated to sequence orders. This could also be interoperability testing of data exchanged across a supply
implemented by an external application. Orders that chain from headquarters to shipping. The model has been
are completed will be deleted while new ones will be integrated with the supply chain simulation developed in
dynamically added to the sequence. The informa- another system (ARENA) using the High Level Architec-
tion to consider in sequencing will be the customer, ture. Production orders and completed order information
estimated lead time of the parts and components, has been exchanged across simulations. Further im-
due-date customer requirements, capacity of the provements in sophistication of the simulation model will
plant, estimated start time for the order, etc. Similar enable it to handle a wider range of products and ex-
sequencing logic will be used before beginning paint change a wider range of messages. It will also largely be
and final assembly processes. run on neutral data, which will be defined outside the
• The delivery of raw materials and parts for assembly model and read in at initialization time. This will reduce
will depend on a forecast of expected production. In the model modification effort. The simulation model will
this case, lead time of the acquisition of parts will be also facilitate the development and testing of a hierarchi-
taken into account. We will also consider the lot siz- cal planning system approach where a central Enterprise
ing of parts during ordering. Requirements Planning (ERP) system controls all opera-
• Initial inventory in the model will either be read into tions including inventory management, products, and
the model or carried over from a previous simulation scheduling. It will also enable the testing for exchange of
run. At the start of the simulation the model need data from the highest level of simulation (supply chain) to
not be empty of parts. the lowest level (machine operation) including Program-
• Additional information will be exchanged between mable Logic Controllers (PLC) systems since generally
the supply chain and assembly plant simulations these would be modeled using different software.
rather than just the orders. As standards for interoperability of information sys-
• The customer order will comprise of more than one tems continue to develop, the need for such generic
product by increasing sophistication in the exchange models to test these standards will increase. The current
and interpretation of messages. project at NIST is aimed at enabling testing of manufac-
• In addition to showing the quantities of parts at the turing engineering processes including design data,
stations, the buffer storage will be made to change engineering, and production planning. When this is facili-
color if the quantities of parts at a station falls below tated it will help reduce costs associated with lack of
a pre-defined level. interoperability of data exchange not only in automotive,
• The logic should allow the plant to stop if the quan- but also in other manufacturing environments.
tities of given parts at a station are used up or if the
quantities have fallen below a predetermined level. ACKNOWLEDGMENT
• There will be a dynamically updated inventory file
to allow sending a message when a replenishment is Work described in this report was sponsored by the Na-
required. tional Institute of Standards and Technology (NIST)
• The model to include taking care of contingencies Manufacturing Interoperability Program, Gaithersburg,
such as breakdowns, downtime, and time to repair. Maryland. The Manufacturing Interoperability Program
supports NIST projects applying information technologies
• Final assembly plant information such as process
and standards-based approaches to manufacturing soft-
times, part routing, and process definition. will be
ware integration problems. The work described was
defined outside the model using XML and read into
funded by the United States Government and is not sub-
the simulation at initialization. The plan is to utilize
ject to copyright.
automotive assembly using the Core Manufacturing
 Kibira and McLean

DISCLAIMER wnloads/oagisaiag/oagisaiag.htm> [Last

accessed on April 16, 2007].
A number of software products are identified in context in Schmenner, R.W. A Machine-Paced Flow Line Process.
this paper. This does not imply a recommendation or en- Production/Operations Management, Science Re-
dorsement of the software products by the authors or search Assosiates, Chicago, IL, 1981, 75-96.
NIST, nor does it imply that such software products are Tan, G., Zhao, N. and Taylor, S.J.E. 2003. Automobile
necessarily the best available for the purpose. Manufacturing Supply Chain Simulation in the
GRIDS Environment, In Proceedings of the 2003
REFERENCES Winter Simulation Conference, ed. S. Chick, P. J.
Sanchez, D. Ferrin, and D.J. Morrice, 1149-1157.
CMSD Product Development Group. 2006. Core Manu- Piscataway, New Jersey: Institute of Electrical and
facturing Simulation Data Information Model (Draft). Electronics Engineers, Inc. .
Part 1: UML Model. NIST. Ulgen, O. and Gunal A. 1998. “Simulation in the Auto-
Goldfarb, C. F. and Prescod, P. 2000. The XML Hand- mobile Industry”, Handbook of Simulation. ed. Jerry
book. 2nd ed. NJ: Prentice Hall PTR. Banks. 547-570, John Wiley & Sons, Inc. .
Hugan, J.C. 2001. Using Simulation to Evaluate Site Traf-
fic at an Automobile Truck Plant. In Proceedings of AUTHOR BIOGRAPHIES
the 2001 Winter Simulation Conference, ed. B.A. Pe-
ters, J.S. Smith, D.J. Medeiros, and M.W. Rohrer, DEOGRATIAS KIBIRA is a Senior Lecturer in the De-
1013 - 1016. Piscataway, New Jersey: Institute of partment of Mechanical Engineering at Makerere
Electrical and Electronics Engineers, Inc. . University in Uganda where he teaches Manufacturing
Jain, S. and Leong, S. 2005. Stress Testing a Supply and Quality systems. He has wide research experience in
Chain Using Simulation. In Proceedings of the 2005 manufacturing simulation and production scheduling. He
Winter Simulation Conference, ed. M.E. Kuhl, N.M. is currently a Guest Researcher at the National Institute of
Steiger, F.B. Armstrong, and J.A. Joines, 1650-1657. Standards and Technology where he is part of the re-
Piscataway, New Jersey: Institute of Electrical and search team involved in the development of a Virtual
Electronics Engineers, Inc. . Manufacturing Environment (VME) for automotive
McLean, C., Riddick, F. and Lee, Y. T. 2005. An Archi- manufacturing. He has a first class honors degree in Me-
tecture and Interfaces for Distributed Manufacturing chanical Engineering from Makerere University and
Simulation. Simulation, Vol. 81, No. 1, 15-32. Simula- Masters and PhD degrees in Manufacturing Engineering
tion Councils Inc. from the University of New South Wales, Australia. His
McLean, C. and Shao, G. 2003. Generic Case Studies for e-mail addresses are <[email protected]>
Manufacturing Simulation Applications. In Proceed- and <[email protected]>.
ings of the 2003 Winter Simulation Conference, ed. S.
Chick, P. J. Sanchez, D. Ferrin, and D. J. Morrice, CHARLES R. MCLEAN is a computer scientist and
1217-1224. Piscataway, New Jersey: Institute of Elec- Group Leader of the Manufacturing Simulation and Mod-
trical and Electronic Engineers, Inc. . eling Group at the National Institute of Standards and
Moon, D. H., Cho, H. I., Kim, H.S., Sunwoo, H. and Technology. He has managed research programs in
Jung, H.Y. 2006. A Case Study of the Body Shop De- manufacturing simulation, engineering tool integration,
sign in an Automotive Factory Using 3D Simulation, product data standards, and manufacturing automation at
International Journal of Production Research, Vol.44, NIST since 1982. He has authored more than 50 techni-
Nos.18-19, 4121-4135. cal papers on topics in these areas. He is on the Executive
Moon, D. H., Song, C. and Ha, J.H. 2005. A Dynamic Board of the Winter Simulation Conference and the Edi-
Algorithm for the Control of Automotive Painted torial Board of the International Journal of Production,
Body Storage, Simulation, Vol. 81, Issue 11, Nov. Planning, and Control. He is formerly the Vice Chairman
2005, 773-787. of the International Federation of Information Processing
National Institute of Standards and Technology (NIST). (IFIP) Working Group on Production Management Sys-
1999. Planning Report 99-1: Interoperability Cost tems (WG 5.7). He holds an M.S. in Information
Analysis of the U.S. Automotive Supply Chain. Engineering from University of Illinois at Chicago and a
Gaithersburg, MD. B.A. from Cornell University. His e-mail address is
OAGi, 2007. Open Application Group’s Integration <[email protected].>
Specification (OASIS) 8.0 SP2 with AIAG Overlay
1.0a. Available on-line via:
<http://www.openapplications.org/do
 Kibira and McLean

Figure 3: Screen shot of the automotive manufacturing assembly plant simulation.

View publication stats