-- An Enabling -- An Enabling Technology for the Technology for the Evaluation of Supply Evaluation of Supply Chains Chains -- Supported by NIST -- Supported by NIST Jeyendran Venkateswaran, Jeyendran Venkateswaran, Mohammed Jafferali, Young-Jun Mohammed Jafferali, Young-Jun Son Son Systems and Industrial Systems and Industrial Engineering Engineering The University of Arizona The University of Arizona (520) 626-9530, (520) 626-9530, [email protected][email protected]
19
Embed
Distributed Simulation -- An Enabling Technology for the Evaluation of Supply Chains -- Supported by NIST Jeyendran Venkateswaran, Mohammed Jafferali,
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Distributed Simulation -- Distributed Simulation -- An Enabling Technology An Enabling Technology
for the Evaluation of for the Evaluation of Supply ChainsSupply Chains
-- Supported by NIST-- Supported by NIST
Jeyendran Venkateswaran, Jeyendran Venkateswaran, Mohammed Jafferali, Young-Jun SonMohammed Jafferali, Young-Jun Son
Systems and Industrial EngineeringSystems and Industrial Engineering
The University of ArizonaThe University of Arizona
Different Types of Different Types of Manufacturing Manufacturing
SimulationsSimulationsEnterpr
iseShop
Equipment
Real World
• Interface with other Interface with other enterpriseenterprise
• Focus of this Focus of this presentationpresentationInterface/Interaction
Big Executor (Shop Level)
ABBrobot
ProlightIBMrobot
FadalCartracAS/RS
TaskOutput Queue
TaskInput Queue
Eshedrobot
Real World or Equipment level simulation
Shop Floor Control Shop Floor Control SimulationSimulation
Shop Level Simulation
Equipment Simulation
Big Executor (Shop Level)
ABBrobot
ProlightIBMrobot FadalCartracAS/RS
TaskOutput Queue
TaskInput Queue
Eshedrobot
Real World or Equipment level simulation
Shop Level Simulation
Equipment Simulation
Evaluation of B2B Evaluation of B2B CollaborationCollaboration
Represented by
Run Time Infrastructure (RTI)
Supplier modelin ARENA(federate)
Distributorin AUTOMOD
(federate)
Manufacturingin PROMODEL
(federate)
NIST AdapterCommon Interface
Common Interface
CommonData Repository
amongCollaborators
NETWORK (LAN or INTERNET)
XML documents, Objects & messages
XML documents, Objects & messages
XML documents, Objects & messages
XML documents, Objects & messages
Represented by Represented by
Requirements to enable Requirements to enable Simulation IntegrationSimulation Integration
• Given multiple simulation modelsGiven multiple simulation models• Requirements (mechanisms) for the Requirements (mechanisms) for the
integration of multiple simulation modelsintegration of multiple simulation models– Time coordination (current simulation time Time coordination (current simulation time
should be the same)should be the same)– Interaction (coordination) or communicationInteraction (coordination) or communication
• Objects (material and information)Objects (material and information)• Messaging (interactions)Messaging (interactions)
• High Level Architecture (HLA) and Run-High Level Architecture (HLA) and Run-Time Infrastructure will provide a basis for Time Infrastructure will provide a basis for simulation systems integrationsimulation systems integration
HLA/RTI to be used directly in HLA/RTI to be used directly in Manufacturing DomainManufacturing Domain
• Is complexIs complex• Requires too much custom coding Requires too much custom coding
from prominent simulation from prominent simulation applications in the manufacturing applications in the manufacturing domaindomain
• So, what about a reusable, easy-to-So, what about a reusable, easy-to-use, and common interface => NIST use, and common interface => NIST AdapterAdapter
Time Management (1)Time Management (1)
• Tnow: current time in a legacy simulation Tnow: current time in a legacy simulation (federate)(federate)– When stand-alone modeWhen stand-alone mode
• Tnow is updated based on original event calendarTnow is updated based on original event calendar
– In federation (distributed simulation)In federation (distributed simulation)• Tnow is constrained by Tnows of other simulations through C and STnow is constrained by Tnows of other simulations through C and S ii
• C: C: current time in distributed simulation current time in distributed simulation (federation)(federation)
• SSii: simulation step size of federate i: simulation step size of federate i– User variableUser variable– Smaller, more tightly coordinated among federatesSmaller, more tightly coordinated among federates
How do we implement the How do we implement the interactioninteraction
and time management?and time management?• Each simulation member can run as stand alone Each simulation member can run as stand alone
How do we model this How do we model this interaction?interaction?
• Each simulation member can run as stand alone Each simulation member can run as stand alone
Information flow is opposite directionfrom the material flow
Functional Specification using Functional Specification using IDEFIDEF
Input
Function
Output
Mechanism
Control
Decomposed FunctionsDecomposed Functions
Decomposed Functions
0 1 open_transaction_as open_transaction_ok_sa
9
2
5 3
order_as$12345$
close_transaction_as close_transaction_ok_sa
O
O
I
I
I
4
T
remove_entity
load_parts_ts
O
I
load_parts_ok_st 8 6 7
generate_transport_order T
transport_order_st
O10
Modeling behavior among members Modeling behavior among members using FSAusing FSA
ImplementationImplementation
• 4 Members have been modeled in Arena4 Members have been modeled in Arena and ProModel and ProModel (AutoMod (AutoMod will be will be included)included)
• Adapter (NIST) has been implemented in Adapter (NIST) has been implemented in COM objectsCOM objects
• Both Arena Both Arena and ProModel and ProModel provides a provides a built-in environment to use COM objects built-in environment to use COM objects as part of their simulation softwareas part of their simulation software
• DemonstrationDemonstration
Performance Test for Performance Test for SuitabilitySuitability
Modeling toolModeling tool SingleSingle Only ArenaOnly Arena
MultipleMultiple Arena, Arena, ProModelProModel
Message sizeMessage size SmallSmall Less than 30Less than 30
LargeLarge More than 300More than 300
Invocation of RTIInvocation of RTI Method 1Method 1
Method 2Method 2
What’s next?What’s next?• Case studiesCase studies• Evaluation scenariosEvaluation scenarios
– Evaluation of virtual enterprises (lead-time, cost, Evaluation of virtual enterprises (lead-time, cost, quality)quality)
• Develop a generic brokerDevelop a generic broker– Information modeling of manufacturing objectsInformation modeling of manufacturing objects– Generic input/output interface from usersGeneric input/output interface from users
• Embedding/adding more scientific Embedding/adding more scientific algorithmsalgorithms– Interaction among members (Petri-net, agent Interaction among members (Petri-net, agent