Scribd
Upload
35 views27 pages

Thomas Peugeot

The document discusses the integration of Model-Based Systems Engineering (MBSE), Product Lifecycle Management (PLM), Mixed Integer Programming (MIP), and robust optimization in the management of the French air defense system, SCCOA. It highlights the complexities of managing a system of systems with numerous stakeholders and the need for effective decision support frameworks. The conclusions emphasize the benefits of using Prolog for communication between tools and the advantages of MIP for optimizing resource constraints and financial costs.

Uploaded by

Moulay Barmak
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
35 views27 pages

Thomas Peugeot

The document discusses the integration of Model-Based Systems Engineering (MBSE), Product Lifecycle Management (PLM), Mixed Integer Programming (MIP), and robust optimization in the management of the French air defense system, SCCOA. It highlights the complexities of managing a system of systems with numerous stakeholders and the need for effective decision support frameworks. The conclusions emphasize the benefits of using Prolog for communication between tools and the advantages of MIP for optimizing resource constraints and financial costs.

Uploaded by

Moulay Barmak
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 27

MBSE, PLM, MIP and Robust Optimization for SoS

management of SCCOA
N. DUPIN (DGA), T. PEUGEOT, C. DUBECQ (MOSS), M.J. SEMBELY
(Airbus)
thomas.peugeot@moss.fr, nicolas.dupin.2006@polytechnique.org
SMQ-MDL-MQAQ-0029-MOSS 1.0
Gentle reminder

System of systems
engineering must be
problem driven*
- How SE methodology is implemented for the complex industrial SoS SCCOA ?
- How to integrate in a tool stack state-of-the-art method of SE and OR to address a
specific problematic ?

* D. Luzeaux, J. Ruault, J. Wippler, Complex Systems and Systems of Systems, John Wiley & Sons (2013)
Agenda

Context presentation

Implementation of SE methods and specific tool stack

RETEX

Conclusion

13 DEC 2016 SC-PPT-OT4001-0436-MOSS 3


SCCOA establishes the Recognized Air Picture 24/7

13 DEC 2016 SC-PPT-OT4001-0436-MOSS 4


SCCOA provides tactical control of military flights

13 DEC 2016 SC-PPT-OT4001-0436-MOSS 5


SCCOA, the French air Defense system, is a real
system of systems

More than 80 systems


(radars, telecom artifacts,
control and command
centers, gateways)

… connected to 80
external systems

More than 150


operational centers

More than 100 sites


SCCOA's management meets Maier’s criteria

A lot of EMA

stakeholders…
SIMMAD DIRISI

… many EMM

programs…

… and many
decisions makers
…

TRS …
TCS

SCCOA Stakeholders
(partial view)

13 DEC 2016 SC-PPT-OT4001-0436-MOSS 7


A permanent issue : SCCOA’s management issues

Systems can be acquired


and/or used independently

SCCOA’s Project Management


(PM) is decomposed into several
management decision levels

SCCOA is interfaced with other


military and civilian programs

Incremental management of
SCCOA to face SoS complexity

13 DEC 2016 SC-PPT-OT4001-0436-MOSS 8


A transient issue : migration of telecom artifacts

Real names have been obfuscated

Issue : no tool is available + schedule uncertainty

13 DEC 2016 SC-PPT-OT4001-0436-MOSS 9


MBSE tooling has made recent progress

*Ernadote, D.: An ontology mindset for system engineering. In: Systems Engineering (ISSE), 2015 IEEE International
Symposium on, pp. 454–460. IEEE (2015)

MBSE enables stakeholders communication with a Domain Specific Ontology on top


of NAF*
System Engineering and Operations Research/Optimization

In common: Decision Support frameworks

SE: structured processes, information and databases.


Questions raised in multi-objective optimization

OR: algorithms and tools for complex decision making

Opportunities to couple SE and OR.

State-of-the-art SE/OR: historic heritage of AI techniques


(sub-gradients, genetics algorithm, Constraint Programming)

Recent helpful advances in OR (generic design of meta-


heuristics, model&run solvers: LocalSolver, MIP, robust
optimization)
13 DEC 2016 SC-PPT-OT4001-0436-MOSS 11
Agenda

Context presentation

Implementation of SE methods and specific tool stack

RETEX

Conclusion

13 DEC 2016 SC-PPT-OT4001-0436-MOSS 12


One tool stack to tackle many issues
The stack enables data reuse across tools

ACCESS MEGA PROLOG

Calendar Logical Deployment Sequencing


Database architecture architecture Tool

Deployment
schedule

Optimized Scheduling
deployment constraints
schedule
Contract Cost &
Robust
Managers Resources
Optimization
Cost Optimization
data
MIP, IBM CPLEX

When a new problem is met, the tool stack is augmented


A Product Lifecycle Management (PLM) tool is
needed to support SoS engineering

System engineering: PLM


Single system
Definition of Definition of spatial &
engineering configurations temporal deployment

System engineering: PLM


System of systems Definition of Definition of spatial &
engineering configurations temporal deployment

Engineering data flow


An ontology for the management of System of
Systems

Main
validation
effort

Main driver is capture of validated models


An pivot ontology for enabling communication
between tools

Defined by
problem of Defined by
interest CM
Communication between tools is enabled by Prolog
and a CM process
Communication between tool is a « change
management » process
Input & output formats are tool neutral (n-
uplets)
Correspondence rules* are expressed with the
Domain Specific Language of the pivot ontology

Knowledge Knowledge
Tool A Base A Base B Tool B
Correspondance
rules*

Change Change Change


Change Change Change Change Change
Change
notice to Control Change to
Request Request to
Change Control Change
notice to
Request
Basetoto
A Request
Base A to
Tool B to
notice Board Request
Base A Request
Base A to Board Tool B to
notice
Tool A Tool A Tool B Tool B
PROLOG

*ISO/IEC/IEEE 42010
Mixed Integer Programming (MIP)/ Robust optimization

PERT scheduling: a polynomial problem


Minimizing total completion time (makespan)
Only precedence constraints among tasks
Implemented in MS Project

Adding resources constraints, RCPSP, a NP Hard problem.

Our objectives:
Minimizing dismantling dates for SCCOA sites
Minimizing financial cost through maintenance costs of obsolete
technologies

Needs for a specific optimization problem


MIP: model&run implementation using OPL Cplex
Robust optimization extension: considering stochastic delays in project
milestones
Plan

Context presentation

Implementation of SE methods and specific tool stack

RETEX

Conclusion

13 DEC 2016 SC-PPT-OT4001-0436-MOSS 19


Lessons learnt using MEGA NAF

MEGA NAF imposed by the DGA for coherence of DGA


architecture modelling.

Automatic generation of chapters of analytical documents


ensures consistent documentation.

Some vigilance points were raised thanks to the NAF SoS


model to detect functional and/or temporal discontinuities.

Generation of a website with selected views and pieces of


information:
clear success amongst stakeholders of SCCOA.
The website gave satisfaction for its ability to capitalize knowledge
amongst stakeholders, it was adopted for the training of
newcomers in SCCOA.
Surprise ! Prolog (b. 1972) match the needs !

Input/output format are


tools neutral

Correspondence rules are


naturally and concisely
written in prolog → Low
maintenance

Prolog is open source,

Having satisfactory
performances with quick
developments encouraged us
to use Prolog.
Lessons learnt using Mixed Integer Programming

IBM Cplex for the MIP model&run computations, OPL script for
data preprocessing.

MIP allows to deal with more complex models: resource


constraints, financial costs and uncertainty with robust optimization
techniques.

For this case, resolution to optimality is easy with Cplex. Other


free and less efficient MIP solvers could be used efficiently.

Generally, MIP requires some specific expertise in modeling to be


efficient, but offers optimality guarantees.

For discrete optimization problems, we recommend for non


experts to use more intuitive modeling frameworks such as
LocalSolver or Constraint Programming.
Plan

Context presentation

Implementation of SE methods and specific tool stack

RETEX

Conclusion

13 DEC 2016 SC-PPT-OT4001-0436-MOSS 23


Conclusions

Methodological conclusions:
NAF extensions : Calendar database, PLM tool with Prolog
Interest of Prolog for concision and maintainability
MIP interest : optimality proof, state-of-the art robust optimization techniques

Industrial conclusions for SCCOA :


Return on investment of the SE methodology
Web site export: a powerful tool for communication amongst stakeholders
and for training of SCCOA newcomers.
Innovating methodology for opportunity questions.
Dismantling old telecom artifacts: planning results presented to the customers
and decision makers
Perspectives

Methodological perspectives:
Generalization of the methodology for other SoS
Excel for pivot files between SE and OR
Using more open source software (COIN-OR)
Multi-objective optimization
Constraint Programming: efficient for scheduling problems, easier for model&run
LocalSolver: large scale model&run optimization solver
Simulation coupled with SE and OR.

Industrial perspectives for SCCOA :


Preparation of next increment SCCOA 5 : architecture design
Other opportunity and scheduling questions
References

System Engineering references :


D. Luzeaux, J. Ruault, J. Wippler, Complex Systems and Systems of
Systems, John Wiley & Sons (2013)
J. Estefan, et al.: Survey of model-based systems engineering (MBSE)
methodologies. Incose MBSE Focus Group 25(8) (2007
Ernadote, D.: An ontology mindset for system engineering. In: IEEE
International Symposium on Systems Engineering (ISSE), pp. 454–460. (2015)

Operations Research references :


D. Bertsimas, M. Sim: The price of robustness. Operations research 52(1),
35–53 (2004)
M. Fischetti, M. Monaci, Light robustness. Lecture Notes in Computer
Science 5868, 61–84 (2009)
J. Schulz, Hybrid solving techniques for project scheduling problems. Ph.D.
thesis (2013)
N. Dupin, Modélisation et résolution de grands problèmes stochastiques
combinatoires: application à la gestion de production d’électricité. Ph.D. thesis,
Univ. Lille 1 (2015)
Appendix: MIP model

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy