Feasibility analysis and efficient routing for a partially automated delivery system within Chalmers campus

Abstract

The use of autonomous driving robot transportation technology to solve the last mile delivery problem is a research hot spot in today’s logistics industry. This thesis aims to analyze the feasibility of applying an Automated Delivery Robot designed by Hugo company to load different sizes of packages and the efficient planning and effectiveness evaluation for the robot. In this thesis, these problems were solved in three steps. First, the GLM model is used to fit the size data of packages shipped by the Chalmers Transportation Center within one year. Secondly, this thesis proposes the concept of package unit to help solve the vehicle routing problem by unifying the volume of the packages. The package unit of each location is calculated through the 3D knapsack problem by the simulated annealing algorithm. Finally, a mixed integer linear programming model was created to optimize the total travel distance of the robot and the related energy consumption was calculated. A case study was conducted by inputting the one day data collected by us into the MILP model, the energy consumption on that day was obtained. The results of the case study shows the use of automated robot for package delivery in university campus is feasible and efficient. This thesis provides suggestions and inspiration for the practical application of automatic transportation on university campus. This thesis also focuses on the energy consumption of automated robots and calculates the approximate energy consumed by automated robots during operation.