OPTIMIZING WASTE MANAGEMENT ROUTES USING HEURISTIC FOR THE VEHICLE ROUTING PROBLEM: A GIS-BASED APPROACH

Abstract

This thesis explores the application of biologically inspired optimization techniques to the problem of urban waste collection routing, focusing on the BLOB algorithm a Physarum-based metaheuristic modeled on the adaptive foraging behavior of Physarum polycephalum. Building on prior research involving Simulated Annealing and Traveling Salesman Problem (TSP) formulations, the work aims to address both the classical TSP and the more complex Vehicle Routing Problem (VRP) under realistic operational constraints. The research is conducted in two phases: (1) benchmarking the performance of the BLOB algorithm against established methods (Simulated Annealing and Lingo Solver) on three representative waste collection routes, and (2) extending the analysis to a full-scale VRP scenario comprising 150 collection points, heterogeneous vehicle capacities, and a centralized depot. Performance is evaluated using metrics such as total distance traveled, execution time, and adherence to capacity limits. GIS-based spatial validation is employed to visualize and substantiate results within the urban context of Tlemcen, Algeria. The findings demonstrate that the BLOB algorithm offers a viable and computationally efficient alternative to conventional heuristics, achieving competitive route optimization while supporting the operational and environmental goals of municipal waste management systems.