PhD Dissertation Proposal by AGRAWAL Pritee

Proactive and Reactive Resource/Task Allocation for Agent Teams in Uncertain Environments

Synergistic interactions between task/resource allocation and multi-agent coordinated planning/assignment exist in many problem domains such as transportation and logistics, disaster rescue, security patrolling, power distribution networks etc. These domains often feature dynamic environments where allocations of tasks/resources may have complex dependencies and agents may leave the team due to unforeseen conditions (e.g., emergency, accident or violation, damage to agent, reconfiguration of environment).

This proposal seeks to provide efficient and scalable mechanisms with abilities to handle dependencies between tasks/resources, non dedication in agent teams and reconfiguration of the environment due to an external event. To that end, we develop generic models to handle task/resource constrained multi-agent planning/assignment for dedicated and non-dedicated agent teams. We design scalable proactive and reactive algorithms that provide provable quality-bounds. The proactive approaches mainly exploit decomposability to solve independent agent planning/assignment problems and provide posterior quality guarantees while the reactive approaches are highly efficient and provide very quick solutions but without quality guarantees. Finally, we empirically show the high scalability and better solution performance of our approaches in comparison with existing work on the real-world and synthetic benchmarks from literature.