Proceedings of the First Workshop

April 13th, 1999 at the University of Salford, UK.

Summary | Papers | Road Map | Discussion Groups

Summary
The workshop was split into three session. In the morning, the paper session gave each participant the opportunity to outline the work of his or her Node. This session revealed the rich diversity of the work in the AI Planning research community and the real requirements of the industrial participants. It was an interactive session with many questions. The on-line proceedings from are set out below (if you would like a hard copy, then please contact the facilitator). In the early afternoon, a brain storming session was held. The session raised many interesting questions that are documented below. The questions revealed a clear split in the perspectives of the participants. There was a group concerned with the requirements of planning applications independent of the technologies applied and a group interested in how AI Planning Research could be better transitioned into practice. The groups were in no way mutually exclusive, but rather different viewpoints on the same questions. The results of each group are also documented below.

Papers

Approximate Reasoning and Artificial Intelligence Group, Antonio Gonzalez, Computer Science and Artificial Intelligence Department, University of Granada, Granada, Spain.
Automatic Knowledge Acquisition Support Tools for Planning, Maria Fox and Derek Long, Department of Computer Science, University of Durham, UK.
Compiling an actions theory to get provably correct operators for planning, Cyril Pain-Barre and Camilla Schwind, Université de la Méditerranée, France.
Knowledge Acquisition Issues in Planning, Lee McCluskey, Department of Computer Science, University of Huddersfield, UK.
Knowledge Acquisition for AI Planning within the O-Plan Project, Austin Tate, Steve Polyak, and Peter Jarvis, Division of Informatics, The University of Edinburgh.
Knowledge, Expertise and Models in AI Planning, Graham Winstanley, University of Brighton.
Machine Learning in Planning, Daniel Borrajo and Ricardo Aler, Adaptive Complex System Laboratory, Universidad Carlos III de Madrid, Spain.
Comment from the Salford Node.
Systematic Construction of Provably Consistent Domain Models for Planning, Susanne Biundo - The University of Ulm, Germany.
Comment from Cogsys Ltd, Simon Chinnick
Comment from BT, Paul Kearney and Simon Thompson.
Comment from Fiat, Nicola Matino.

Road Map

The following road map was produced during this session:

Knowledge Acquisition Issues

Research Questions

Research Products

The Nature of Knowledge Acquisition AGENTS INVOLVED IN THE KA PROCESS  
Constitution of Knowledge Acquisition Methods LEARN FROM OTHER AREAS, FORMAL METHODS, EVALUATION OF KA METHODS, TOOL-SUPPORT, ROLE-OF-LEARNING, ROLE-OF-LIBRARIES.  
Knowledge Requirements OVERALL REQUIREMENTS, ACTION KNOWLEDGE, HUMAN PLANNING KNOWLEDGE, SEARCH CONTROL.  
Selecting Planning Techniques for a given Domain TAXONOMY OF PLANNING METHODS,
TAXONOMY OF PLANNING DOMAINS, TECHNIQUE SELECTION.
 
Validation, Verification, and Maintenance of Domain Models VALIDATION AND VERIFICATION, MAINTENANCE, DYNAMIC DOMAINS.  
Knowledge Rich Solutions TRADE-OFF.  

Discussion Groups