Distributed and Multi-Agent Planning: Challenges and Open Issues

Planning is a well known and studied field of Artificial Intelligence. Multi-Agent Planning concerns the construction of plans for a group of autonomous agents that can interact. The aim of multi-agent planning is to automatically find a solution such that, if every agent executes successfully his plan, the environment changes to a goal state. The solution can be found either by centralized or distributed algorithms. In this work, we survey recent contributions in the fields of distributed and multi-agent planning. We define the problem, briefly outline possible different classifications of the multi-agent planning problem and present the state of the art in this field. Finally, we report open challenges and issues that may be addressed in the following years.