Working Session I: Saturday Afternoon
 

Group One: Engr II, Conference Room 2301 (MEE Dept.) 3:30-5:00 PM

Facilitators: Hassan Khalil & Le Yi Wang

Participants: Ackermann, Alleyne, Bailey, Bamieh, M.F. Chang, S. Chin, Cikanek, Daleh, Diop, Freudenberg, Yanakiev

Working Session Topics and Issues: Current Practices  How does one derive models for control analysis, design, optimization and verification?  How to choose the appropriate analysis/control tool for a particular system given an existing modeling approach?  How to incorporate the modeling, analysis and control developments at the initiation of a problem? - An analogy would be Design for Manufacturing.

 
 

Group Two: Engr II, Class Room 3301 (ChE Dept.) 3:30-5:00 PM

Facilitators: Masayoshi Tomizuka & Takehiko Fujioka

Participants: J. Cook, C. Gerdes, L. Gielmo, Guzella, Hedrick, Hofbauer, Ioannou, Kalkkuhl, Kokotovic, Kolmanovsky

Working Session Topics and Issues: Feasibility of Formal Integration  Is it possible to integrate modeling and control in a unified analytical framework or is it an "art" best left to the experienced control designer?  If it is possible, what is needed from the research community to seamlessly integrate modeling, analysis and control?  Do System Theoretic tools exist? - If so, which ones? What are their advantages and drawbacks?  - If not, what needs to be developed?

 
 

Group Three: MRL Bldg., Conference Room 2053 (MRL) 3:30-5:00 PM

Facilitators: Keith Glover & Jessy Grizzle

Participants: Horowitz, Kotwick, Krogh, Lukich, Milot, Mezic, B. Paden, H. Peng, B. Powell, I. Rhodes, Winkelman

Working Session Topics and Issues: Extracting insight from examples  Are there commonalties to the successful integration of modeling, analysis, and control?  Based on case studies presented here in this workshop, and other information at the disposal of the workshop participants, identify what some of these are.  What are the commonalties for the cases where integration of modeling, analysis, and control has failed? -One example would be where the model is developed by one community (e.g. physicists) and the control is attempted by another.

 

Group Four: Engr III, Conference Room 116 (Materials Dept.) 3:30-5:00 PM  Facilitators: J.D. Powell & S. Sivashankar Participants: Rizzo, Rizzoni, R. Shankar, R. Smith, Stefanopoulou, J. Sun, Tabaczynski, A. Teel, Tsao, Tsugawa, von Niewstadt Working Session Topic: Bringing Integration to the Classroom.  What is the most appropriate and effective way to teach an integrated approach to Modeling and Control?  Change in pedagogy?  Interdisciplinary teams?  Project oriented classes instead of exam oriented classes?  "Canned" labs vs. "open ended" projects?

 

Group One Engr II, Conference Room 2301 (MEE Dept.) 2:00-3:00 PM

Facilitators: J. Karl Hedrick & Anna Stefanopoulou

Participants: Ford, von Nieustandt, Glielmo, Kolmanovsky, Tabaczynski, Teel, Kotwick, Wang, Yanakiev, Smith

Working Session Topics and Issues: Current Practices  How does one derive models for control analysis, design, optimization and verification? How to choose the appropriate analysis/control tool for a particular system given an existing modeling approach? How to incorporate the modeling, analysis and control developments at the initiation of a problem? - An analogy would be Design for Manufacturing.

 
 
 

Group Two: Engr II, Class Room 3301 (ChE Dept.) 2:00-3:00 PM

Facilitators: Jurgen Ackermann & Petros Ioannou

Participants: Chang, Rizzoni, Bamieh, Freudenberg, Ackermann, J.D. Powell, Dahleh, Meinhart, Kalkkuhl, B. Powell, Mezic, Lukich.

Working Session Topics and Issues: Feasibility of Formal Integration  Is it possible to integrate modeling and control in a unified analytical framework or is it an "art" best left to the experienced control designer? If it is possible, what is needed from the research community to seamlessly integrate modeling, analysis and control? Do System Theoretic tools exist? - If so, which ones? What are their advantages and drawbacks?  - If not, what needs to be developed?

 
 
 

Group Three: MRL Bldg., Conference Room 2053 (MRL) 2:00-3:00 PM

Facilitators: Gianfranco Rizzo & J. Chris Gerdes

Participants: Chin, Fujioka, Tomizuka, Rizzo, Sivashankar, Gerdes, Seaborg, Diop, Shankar, Tsao, Bailey, Glover.

Working Session Topics and Issues: Extracting insight from examples  Are there commonalties to the successful integration of modeling, analysis, and control? Based on case studies presented here in this workshop, and other information at the disposal of the workshop participants, identify what some of these are. What are the commonalties for the cases where integration of modeling, analysis, and control has failed? -One example would be where the model is developed by one community (e.g. physicists) and the control is attempted by another.

 
 
 

Group Four: Engr III, Conference Room 116 (Materials Dept.) 2:00-3:00 PM

Facilitators: Jim Winkelman & Huei Peng

Participants: Cook, Guzella, Grizzle, Hofbauer, Sun, Rhodes, Milot, Horowitz, Cikanek, Ponti, Alleyne, Paden, Tsugawa, Kokotovic, Krogh.

Working Session Topic: Complexity and Scalability  Can the chosen integration approach scale with problem complexity/detail or should the approach change with problem complexity/detail E.g. traffic flow systems vs. individual vehicle systems. E.g. chemical combustion vs. I/O engine performance. Is there a breakdown by either the modeling or the controller (or both) that limits or prohibits an integration as the problem becomes scaled to larger size of finer resolution? If so, what?s more likely to ?fail?: the model representations or the control approaches?

 
 
 
 

 

Group One: Engr II, Conference Room 2301 (MEE Dept.) 3:15-4:15 PM

Facilitators: Andrew Alleyne & Keith Glover

Participants: Bamieh, B. Powell, Chang, Grizzle, Bailey, Hofbauer, Seaborg, Guzella, Tomizuka, Ioannou, Ford, Lukich, Tabaczynski

Working Session Topic: Problem Classification & Industrial Needs  Can models and modeling techniques be categorized along with control techniques to give a ?look-up? table approach to integrated problem solutions? If so, what. If not, why not. What are industrial needs (wants) for an integration of modeling and control? Is there a need or can most problems be solved efficiently without it?

 
 
 
 

Group Two: Engr II, Class Room 3301 (ChE Dept.) 3:15-4:15 PM

Facilitators: Bruce Krogh and Roy Smith

Participants: Sivashankar, Paden, Fujioka, Peng, Chin, Dahleh, Gerdes, Yanakiev, Wang, Rizzoni, Kotwick.

Working Session Topics and Issues: Feasibility of Formal Integration  Is it possible to integrate modeling and control in a unified analytical framework or is it an "art" best left to the experienced control designer? If it is possible, what is needed from the research community to seamlessly integrate modeling, analysis and control? Do System Theoretic tools exist? - If so, which ones? What are their advantages and drawbacks?  - If not, what needs to be developed?

 
 
 

Group Three: MRL Bldg., Conference Room 2053 (MRL) 3:15-4:15 PM

Facilitators: Lino Guzzella and TC Tsao

Participants: Rizzo, Cikanek, Teel, Kalkkuhl, Cook, Hedrick, Ponti, Winkelman, J.D. Powell, Freudenberg, Rhodes, Ackermann, Sun

Working Session Topic: Bringing Integration to the Classroom  What is the most appropriate and effective way to teach an integrated approach? Change in pedagogy? Interdisciplinary teams? Project oriented classes instead of exam oriented classes? "Canned" labs vs. "open ended" projects?

 

 
 

Group Four: Engr III, Conference Room 116 (Materials Dept.) 3:15-4:15 PM

Facilitators: Sette Diop & Ilya Kolmanovsky

Participants: Horowitz, Mezic, Kokotovic, Meinhart, Stefanopoulou, Tsugawa, von Nieustandt, Glielmo, Milot, Shankar.

Working Session Topic: Complexity and Scalability  Can the chosen integration approach scale with problem complexity/detail or should the approach change with problem complexity/detail E.g. traffic flow systems vs. individual vehicle systems. E.g. chemical combustion vs. I/O engine performance. Is there a breakdown by either the modeling or the controller (or both) that limits or prohibits an integration as the problem becomes scaled to larger size of finer resolution? If so, what?s more likely to ?fail?: the model representations or the control approaches?

Thank you for agreeing to facilitate a working session. As a facilitator, you should feel free to do whatever you feel is best. However, you may want to keep the following in mind:

1. Your first task is to decide which of the two facilitators will be the discussion moderator and which will be the recorder. The recorder will write key concepts and points to assist in the presentation of the working session?s findings on Sunday. The purpose of dividing the tasks is to allow as free a flow of discussion as possible.
2. Make sure the discussion is not dominated by a few people, but involves everyone in the group.
3. Try to keep the discussion on topic. You may want to prepare a few of your own questions or points of discussion before your working session in that regard.
4. At the end of your working session, the two facilitators should get together with counterparts from other groups who were working on the same topic, if there were any. You should then

• Select a speaker for the summary session on Sunday afternoon.
• Discuss the content of the presentation to reflect the comments and consensus of each group involved. If there are strong opinions that were not a part of the group consensus these must be included as well as they are a valuable source of information.

1. On Sunday, the selected speaker will make a five minute presentation on the topic including: issues identified, recommendations and/or conclusions, etc.

(1) Working Session Topic:

• How does one derive models for control analysis, design, optimization and verification?
• How to choose the appropriate analysis/control tool for a particular system given an existing modeling approach?
• How to incorporate the modeling, analysis and control developments at the initiation of a problem?

• Is it possible to integrate modeling and control in a unified analytical framework or is it an "art" best left to the experienced control designer?
• What is needed from the research community to seamlessly integrate modeling, analysis and control?
• Do System Theoretic tools exist?

• Are there commonalties to the successful integration of modeling, analysis, and control?
• Based on case studies presented here in this workshop, and other information at the disposal of the workshop participants, identify what some of these are.
• What are the commonalties for the cases where integration of modeling, analysis, and control has failed?

(4) Working Session Topic:

• What is the most appropriate and effective way to teach an integrated approach?
• Change in pedagogy?
• Interdisciplinary teams?
• Project oriented classes instead of exam oriented classes?
• "Canned" labs vs. "open ended" projects?

(5) Working Session Topic:

• Can models and modeling techniques be categorized along with control techniques to give a ?look-up? table approach to integrated problem solutions?
• If so, what. If not, why not.
• What are industrial needs (wants) for an integration of modeling and control?
• Is there a need or can most problems be solved efficiently without it?

(6) Working Session Topic:

• Should the chosen approach scale with problem complexity/detail or should the approach change with problem complexity/detail
• E.g. traffic flow systems vs. individual vehicle systems.
• E.g. chemical combustion vs. I/O engine performance.