Planning with Complex Data Types in PDDL
Practically all of the planning research is limited to states represented in terms of Boolean and numeric state variables. Many practical problems, for example, planning inside complex software systems, require far more complex data types, and even real-world planning in many cases requires concepts such as sets of objects, which are not convenient to express in modeling languages with scalar types only. In this work, we investigate a modeling language for complex software systems, which supports complex data types such as sets, arrays, records, and unions. We give a reduction of a broad range of complex data types and their operations to Boolean logic, and then map this representation further to PDDL to be used with domain-independent PDDL planners. We evaluate the practicality of this approach, and provide solutions to some of the issues that arise in the PDDL translation.