Optimization and optimal distribution network proposal of waste pick-up routes in BRANTNER GmbH

This thesis addresses the pressing issue of waste collection route optimization for a subsidiary of the Austrian company Brantner, which offers a broadening portfolio of sustainable waste management solutions across Slovakia. Together with Brantner’s growing innovations and expanding operational areas, the complexity follows, necessitating optimization. Thus, the primary objective of this thesis is to identify key operational inefficiencies in the waste collection pickup-routes of Brantner and in turn optimize the majority of these inefficiencies on a subset of the routes. Initially, the thesis introduces the concepts of mathematical optimization and the optimization problem, while delving deeper into specific types of optimization problem methods and algorithms, such as the Heuristics and Metaheuristics methods, or the Ruin and Recreate algorithm. Within the practical part of this thesis, Brantner’s waste collection routes are identified to be a CHVRPTW-LR optimization problem type, which is a specific variant derived from the original VRP. Moreover, a detailed analysis was performed, using a prominent analytical and visualization tool, that uncovered key insights into Brantner’s waste collection operations. A part of this analysis was also dedicated to highlighting the crucial role that data quality holds in generating added value insights and feasible optimization solutions. With the provided analytics, the thesis was able to discover many data discrepancies, waste collection inefficiencies and bottlenecks that were then addressed by the optimization. For the optimization, the thesis used a specialized DW software that employs a meta-heuristic Ruin & Recreate algorithm with the use of the Jsprit open-source tool. The optimization showed very promising results, while it also had its limitations which were caused by the optimization problem simplifications and poor data quality. Overall the approach applied in this thesis has demonstrated a great potential for the improvement of Brantner’s waste collection pickup-routes, as well as the optimization of similar routing problems. The findings, show that 65% of the waste collection transit time, 63% of the vehicle-related costs, and 64% of the total distance travelled have the potential to be reduced, generating a promissing saving potential. Thus, concluding that the hypothesis expectation of this thesis was met considerably well (almost three fold).