Iterative Constraint-based Repair for Multiagent Scheduling

We propose a new integrated architecture for distributed planning and scheduling that exploits constraints for problem decomposition and coordination. Our goal is to develop an efficient method to solve densely constrained planning / scheduling problems in a distributed manner without sacrificing solution quality. We implemented a prototype system, called CAMPS, in which hierarchy of intelligent agents try to coordinate their actions for "satisficing" planning / scheduling results by handling several intraand inter-agent constraints. In this paper, we show the repair-based methodology for distributed planning / scheduling and the constraint-based mechanism for dynamic coalition formation among agents.