Waste routing optimization with variable network structure with smart waste containers

Abstract

As long as humanity exists, it will continue to produce waste. For a livable society, these wastes need to be collected and disposed of at regular intervals. The collection and disposal of these wastes involve many processes and can be quite costly. This study aims to optimize the journey of solid waste from generation to a disposal facility with minimum time and cost, and the solution of the Capacity-constrained Vehicle Routing Problem (CVRP) is used as an optimization technique in this study.A dynamic perspective was added to the capacitated CVRP by incorporating a decision mechanism that determines whether containers should be collected or not based on their predetermined fill levels. The Internet of Things methodology was utilized, and fill sensors were added to the containers. Genetic algorithms, one of the metaheuristic methods, were used in the algorithm's solution, which was implemented through a Python program. After testing the algorithm on a trial network, a specific location was selected. The chosen region is Gerzele Neighborhood in Denizli Province, and the TURKSTAT data pertaining to the neighborhood were used in the calculations. The coordinate data for the neighborhood and the locations of the containers were mapped, and transportation data was generated by integrating them into the Visum program. The collection process was carried out based on the fill levels of the containers using the sensors, and several scenarios were tested. It was determined that cost, distance traveled, and collection times for waste logistics could be reduced by 20-40% based on the obtained results.