AI quality control in competitive recycling facing material contamination

In a typical material recycling supply chain, material recovery facilities (MRFs) are generally classified as dirty MRFs(dMRFs) whose material faces contamination and clean MRFs(cMRFs) whose material is of high clarity. As the downstream manufacturer, purchasing from dMRFs or cMRFs faces the trade-off between material purchasing price (dMRF's material is cheaper) and the waste disposal cost (dMRF's contaminated material will be wasted and disposed of). This also motivates dMRFs to adopt an AI quality control system to eliminate contamination. We build a game-theoretic model to analyze the decision-makers’ incentive of AI adoption and interesting findings include: (1) Even AI quality control pushes the purchasing price upward, dMRF's supply quantity can be surprisingly reduced, indicating “inefficient use of AI”; (2) AI quality control cost and the manufacturer's waste disposal cost exhibit a substitutable relationship in promoting the dMRF's AI adoption. These findings provide important insights for managers in formulating purchasing strategies, investment decisions, and AI adoption. They are suggested to pay attention to win-win-win situations regarding the dMRF's profit, the manufacturer's profit, and the system's environmental sustainability with the dMRF's AI quality control. Since all-win situations for the stakeholders will not sustain as the equilibrium, subsidy schemes should be designed to improve the cMRF's profit.