Track:
Contents
Downloads:
Abstract:
We outline a approach for closely integrating classical automated planning with scheduling in a manner designed to maximize the opportunity for leveraging the strengths of each. A fundamental capability needed to implement the system is an efficient resource-lifted planning graph, and we introduce the process for achieving it here. We report experimental evidence that the resource-abstracted model functions well and provides an effective means of constraining the entailed search space via propagation through the planning graph of the set of valid resource instances for each proposition.