International Workshop on

Planning under Uncertainty and Execution Control
for Autonomous Systems

PROGRAM


[ Local Information ]

[ Venue ]

[ Hotel ]


Overview ]


Scope ]


Format ]


Submissions ]


Important Dates ]


Organisation ]

June 6-10th, 2006
The English Lake District, Cumbria, United Kingdomn


Held in conjunction with

ICAPS'06
16th International Conference on Automated Planning & Scheduling

 

PROGRAM


Overview

The variety of autonomous systems is increasing both in industry and academia (AUV, UGV, UAV, Robots, Space Probes), which addresses a broad range of missions (space exploration, search and rescue, defense and security,...). Such systems must operate with limited human intervention in a changing environment and behave efficiently in spite of mission updates, failure recovery or resource scheduling concerns. Recent experiences show that simple planning and execution control techniques may endanger the system when uncertain information or limited knowledge about the environment is available. In order to achieve a consistent or optimized behavior, planning and execution control problems, addressed separately in the past, have to be solved in a combined and coordinated way. The environment being partially known by an autonomous system, plan generation and execution must consider uncertain information, biased data, unpredicted or contingent events. Moreover, planning and execution control have to satisfy different time scales, depending on mission, system and environment tempos. Lastly, uncertainty may result from approximations at the planning level or unexpected events may occur due to a lack of control. Various theoretical and practical aspects of planning and execution control have been studied so far (Constraint solving, combinatorial optimization, Markov decision processes, controllable networks, Petri nets...). However, a broad range of hybrid algorithms, architecture designs and alternative approaches may be explored during the workshop.

Back to top ]

Scope

This multi-disciplinary workshop invites submissions from researchers in industry and academia. Submissions on planning and execution control techniques as well as practical experiences with experimental or deployed systems (including undersea, ground and aerospace systems ...) are particularly encouraged. Relevant topics include but are not limited to:

  • Planning, scheduling and execution control techniques with uncertain, incomplete or biased data, temporal uncertainty, spatial uncertainty or geolocalisation issues. Robust planning, scheduling and execution control.
  • Uncertainty and constraints in AI planning, constraint based models, constraint solving for planning under uncertainty, constrained optimization techniques and specific consistency management methods (model decomposition, propagation algorithms, arc-consistency...).
  • Search techniques for planning under uncertainty, including anytime, online and continuous search; generic versus domain specific heuristics; hybrid search techniques (local search, branch and bound, binary searches, greedy algorithms, global search, repair, ...). Time-constrained and time-dependent search techniques.
  • Hybridized planning approaches: planning and scheduling, planning and execution, scheduling and execution monitoring, energy-aware planning; planning and control methods, including model-predictive or adaptive control.
  • Planning with domain constraints of activities (land, air, undersea, space and other systems of interest with autonomous behaviors). Real-world applications in various domains, including undersea, land, air and space.
  • Decision theoretic planning with uncertainty for autonomous systems. Probabilistic approaches, Markov Decision Processes, Modeling with Bayesian networks.
  • Planning in multi-agent systems, including planning for cooperative/collaborative agents; plan coordination under uncertainty; and robust distributed planning / execution control / scheduling.

Back to top ]

Format

The workshop format will consist of technical presentations based on accepted papers, followed by a discussion session on each paper, led by an assigned moderator. As for "Planning Under Uncertainty for Autonomous Systems (2005)", a commentary process will be implemented inspired from the "Int. Workshop on P&S of Space (2000)". This process allows authors to have relevant feedback on the paper. This is also an excellent way to initiate discussions between author and attendees. Position papers will be briefly presented.

Back to top ]

Submissions

Submissions may be regular papers (10 pages) or short position papers (2 pages). Position papers should focus on relevance to application domain, positioning, and innovative ideas. For all papers, the format should conform to the AAAI style template. Papers will be reviewed by at least two referees.

Submissions, in PDF format only, should be sent by email to mailto:christophe.guettier@sagem.com using the subject line ICAPS'06 Workshop Submission.

All workshop attendees must be registered for ICAPS'06.

Back to top ]

Important Dates

The schedule of important dates for the workshop is as follows:

Paper submission deadline 14 March 2006
Notification of acceptance 7 Avril 2006
Camera-ready version deadline 18 April 2006 (hard)
Workshop date 6-7 June 2006

Back to top ]

Organisation

Organising Committee

Christophe Guettier (program chair)
SAFRAN SA
Paris, France
Email: mailto:christophe.guettier@sagem.com

Programme Committee

Roy Turner (University of Maine - USA)
Patrick Fabiani (ONERA, France)
Felix Ingrand (LAAS/CNRS, France)
Mark Boddy (Adventium Labs - USA)
Wheeler Ruml (Palo Alto Research Center, USA)
David Ferguson (Canergie Mellon University - USA)
Gerard Verfaillie (ONERA, France)
Eric Jacopin (CREC Saint-Cyr - France)
Neil Yorke Smith (SRI International, USA)
Mark Wallace (Monash University - Australia)
Abdel-Illah Mouaddib (Universite de Caen - France)
Alessandro Saffiotti (Orebro University - Sweden)
Mark Wallace (Monash University - Australia)
Russel Knight (Jet Propulsion Lab - USA)
Nicolas Meuleau (NASA Ames Research Center - USA)

Back to top ]