Addressing the ‘last mile’ deficiency: A bi-objective model for e-waste reverse logistics network design

Although China’s official e-waste recycling rate has been higher than the global average over the past decade, the certified reverse logistics network, primarily composed of licensed disassemblers, faces fierce competition from uncertified channels due to limited community access to the certified network. This paper develops a robust bi-objective mixed-integer linear programming model to help certified network decision-makers establish community collection and treatment centers that address the ‘last mile’ deficiency, aiming to provide accessible services to communities competing with uncertified channels while maintaining economic viability. The robust bi-objective mixed-integer linear programming model considers maximizing the collection rate and the profit of the three-echelon reverse logistics network. A piecewise function is proposed to simulate competition between community collection and treatment centers and uncertified channels based on prior literature. Two realistic uncertainty parameters—the lower boundary of the entire collection and the reuse rate—are incorporated into the model, which deliberately addresses the dynamics of collection and processing complexity. A real-world case is presented to validate the model's effectiveness in supporting location decision-making for community collection and treatment centers. The Gurobi optimizer is used to solve the robust model, and the resulting optimal solutions are presented on the Pareto front. ArcGIS software illustrates the eight selected community collection and treatment centers, which achieved a 61.6 % collection rate and a profit of 0.15 million RMB, as shown on the map. This study presents a reliable and effective model for location decisions for community collection and treatment centers under realistic uncertainties, facilitating licensed disassemblers’ decision-making that balances regulatory and economic objectives.