Learning-embedded disassembly petri net for process planning

The growing concerns for material resources, energy conservation and landfill capacity have put much pressure on manufacturers, charging them with the responsibility for their outdated products. However, obstacles arise when introducing product/material recovery in the economic landscape due to much uncertainty inherent in the process (e.g., prevailing condition of reclaimed products and the level of human intervention). This paper presents a rigorous model that accounts for such system dynamics in disassembly process planning (DPP), a critical stage to the efficiency of product/material recovery. In particular, this model with the learning capability will be able to: (1) mathematically represent the operational planning of disassembly in the light of uncertainty (i.e., the quality of reclaimed products and the impact of human intervention); (2) accumulate and exploit "knowledge" of system performance via the observation of the process behavior; and (3) dynamically derive a cost-effective disassembly plan.