Case Study: Collective Resource Reasoning

As a case study of the application of constraint-satisfaction representation, in this chapter we consider how to extend a traditional backward-chaining, partial-order planner to one that could reason about resources. We present an extended-plan language in which each variable is associated with a finite set of domain values. We discuss the associated changes this brings to planning algorithms. In addition, we present empirical results to demonstrate the efficiency gain of the extended planner, and to answer some of the utility questions regarding planning with constraint satisfaction