Abstract:

Efficient and sustainable municipal waste collection is a critical challenge in urban management, requiring data-driven strategies that align operational performance with environmental objectives. This paper presents an illustrative example of applying an approach that combines clustering algorithms, a core technique in artificial intelligence, with mixed-integer linear programming (MILP) to optimize the division of a city into sustainable solid waste collection sectors. Using real-world data from Tarn´ow, Poland, we apply k-means algorithms to generate spatially and quantitatively balanced clusters of solid waste collection points. These clusters were used as the basis for essential parameters of a square grid used to solve the Sustainabling Sectorization of Municipal Solid Waste Collection Problem using an MILP model designed to minimize either disparities in route lengths or collected solid waste volumes across a heterogeneous vehicle fleet,
including diesel and electric trucks. The proposed approach supports strategic decision-making by enabling urban planners and waste management authorities to evaluate alternative sector layouts, fleet configurations, and environmental trade-offs. The newly developed optimization problem contributes to developing intelligent, adaptable tools for sustainable city logistics. The methodology is implemented in open-source software and is transferable to other municipalities facing similar planning challenges. The results of computational experiments are presented, and practical use is demonstrated.