Schedule for ICAPS 2005 workshop on Multiagent Planning and Scheduling
 

June 6, 2005

text only

In the schedule below are listed questions for open discussion throughout the day. Please send Brad Clement (bclement@jpl.nasa.gov) other ideas for discussion topics or feedback on these. All participants, feel free to prepare responses before or during the workshop. You can send me powerpoint or text, and you can present it during the discussion sessions. Just keep these discussion presentations short (1-2 minutes) since we don't have much time. If my note-taking skills are good enough, I will try to post responses on the web page afterwards.

When:	What:	Who:
8:30 - 8:40	Opening remarks	Brad Clement
8:40 - 9:30	Invited talk - A Spectrum of Strategies for Multiagent Plan Coordination	Ed Durfee
9:30 - 10:30	Interleaved planning and execution Honeywell's COORDINATORs Project Coordinating Agile Systems Through the Model-based Execution of Temporal Plans Execution Monitoring and Replanning with Incremental and Collaborative Scheduling	D. Musliner, J. Phelps T Léauté, B Williams D. Wilkins, S. Smith, L. Kramer, T. Lee, T. Rauenbusch
10:30 - 11:00	Coffee Break
11:00 - 11:15	Discussion: Where can multiagent planning make significant impact in application? How is coordinating humans different than coordinating agents? Are there techniques that are more appropriate for one than the other?
11:15 - 12:15	Evaluation & Market-based approaches Determining Task Valuations for Task Allocation Computing the Communication Costs of Item Allocation Self-interested Planning Agents using Plan Repair	D. Han, K. S. Barber T. Rauenbusch, S. Shieber, B. Grosz R. van der Krogt, M. deWeerdt
12:15 - 12:30	Discussion: Auctions (market mechanisms) are almost always used for assigning tasks for execution or allocating resources in a multiagent system, but this is divorced from the actual planning algorithm. How can agents plan for auctions or use auctions to plan? Should they? How far have we gotten, and how far do we need to go in evaluating multiagent planning (differently than single agent planning)? Should there be benchmarks or competitions?
12:30 - 2:00	Lunch Break
2:00 - 2:50	Invited Talk - Resolving Conflicts about Teamwork: Hybrids to the rescue	Milind Tambe
2:50 - 3:30	POMDPs Bounded Policy Iteration for Decentralized POMDPs Exploiting Interaction Structure in Networked Distributed POMDPs	D. Bernstein, E. Hansen, S. Zilberstein R. Nair, P. Varakantham, M. Tambe, M. Yokoo
3:30 - 4:00	Coffee Break
4:00 - 4:15	Discussion: Are POMDP representations limiting? If so, how? Do existing approaches apply to self-interested agents (stochastic games)?
4:15 - 5:35	Representations & temporal constraints From Multiagent Plan to Individual Agent Plans Planning for Multiagent Environments: From Individual Perceptions to Coordinated Execution Robust Distributed Coordination of Heterogeneous Robots through Temporal Plan Networks Managing Communication Limitations in Partially Controllable Multi-Agent Plans ASET: a Multi-Agent Planning Language with Nondeterministic Durative Tasks for BDD-Based Fault Tolerant Planning	O. Bonnet-Torrès, C. Tessier M. Brenner A. Wehowsky, S. Block, B. Williams J. Stedl, B. Williams R. Jensen, M. Veloso
5:35 - 6:00	Discussion: How is centralized planning for multiple agents different than planning for concurrent action? Should this still be called multiagent planning? What research questions/challenges are being ignored in multiagent planning?