Design and implementation of an adaptive process planner for disassembly processes

This paper presents a methodology for designing and implementing disassembly plans and processes using Petri nets (PN's) for remanufacturing obsolete products. Process modeling and planning using PN's are discussed. The proposed PN's accommodate product topology and precedence relations during disassembly. Decision values are attached to PN's to enable consideration of uncertainty during process execution. The resulting planner guarantees the plan's optimality when product remanufacturing parameters are known. In cases of uncertainty about product parameters, an adaptation scheme is designed and implemented to facilitate the goal of maximum benefit. Three planning and execution algorithms (globally fixed, limited adaptive, adaptive) are presented for disassembly processes. They are demonstrated comprehensively through a robotic disassembly system, and its use is demonstrated in disassembly of a car radio.