Hybrid Fleet Optimization for Waste Collection: Addressing Urban Constraints Using OR-Tools

Abstract

Heuristic and metaheuristic algorithms have demonstrated potential applicability to optimize waste collection in the last few years. Both algorithm classes are able to provide an optimized solution to perform the waste collection task, minimizing the total distance traveled and costs in the process. However, the optimization is often explored with no regard to the reduced truck mobility in some regions. In Europe, most cities have neighborhoods with narrow streets where trucks have trouble moving around, making the waste collection process complicated. In this regard, this paper presents the results obtained for the selective waste collection (paper, glass, and plastic waste) optimization in constrained urban spaces. By comparing the heuristic and metaheuristic algorithms available in the open-source OR tools in systems comprised of 50–100 dumpsters, the best algorithm was selected and further used to study waste collection optimization in a constrained urban space. For this purpose, a hybrid fleet comprised of regular-size combustion trucks and small-size electric vehicles was considered. The detailed analysis provided in this paper regarding the influence of system size at the optimization level using the selected heuristic and metaheuristic algorithms provides insights into a wide range of waste collection scenarios. In addition, the case study demonstrated that a hybrid fleet was able to reduce the CO2 emission by 31.9% compared to a regular fleet.