Task decomposition and analysis of robotic assembly task plans using Petri nets

This paper describes a framework for robotic task sequence planning which decomposes tasks into operations sequences for a generic robotic workcell. The approach provides robust execution of tasks through properties of: traceability-implicit mapping operations representation, and viability-retaining multiple execution. Given the descriptions of the objects in the system and all feasible geometric configurations and relationships among these objects, an AND/OR net which describes the relationships of all feasible geometric states and associated feasibility criteria for net transitions is generated. This AND/OR net is mapped into a Petri net which incorporates all feasible sequences of high level operations. The resulting Petri net is then decomposed in a stepwise manner into lower level Petri nets of which each transition can be directly implemented by control commands or command sequences based on devices and objects in the system, or, by lower level planning transitions corresponding to path planning, grasp planning, fine motion planning, etc. The property analysis for different levels of decomposition is also presented, and the inheritance of properties between levels is defined. All possible task sequences could be found using a search algorithm based on feasible system states. The shortest sequence may be chosen from the lowest level decomposition and is guaranteed to be the optimal output of the hierarchical planning system to efficiently implement the desired tasks