Local Search Heuristsics for Generative Planning

Local-search approaches promise very interesting features for planning, such as an anytime computation and an uncomplicated handling of the environment's dynamics. However, most local-search approaches require the availability of maximal structures what causes problems of scalability. Local search within a generative planning process is rarely looked into. This article presents some basic heuristics for generative planning, and demonstrates the approaches' capabilities by way of solving variations of combined planning and scheduling in the Logistics