Collaborative Gain Assessment for the Dynamic Multi-Objective Vehicle Routing Problem

Abstract

Inefficiency in trucking has both economical and ecological consequences. In literature, horizontal collaboration is identified as one of the most promising solutions to this problem. However, collaboration is not often encountered in trucking industry, since carriers are hesitant to join collaborations. This is caused by a fear of sharing information and an uneven distribution of requests between carriers. Furthermore, the amount of collaborative gain remains still unclear when dealing with a dynamically changing environment, which is often present. This work aims to fill this knowledge gap. Firstly by developing a vehicle routing problem (VRP) model, which includes dynamic requests, horizontal collaboration and customer satisfaction. Next to that, a solution method is developed which is based on Adaptive Large Neighborhood Search (ALNS). The performance of the implemented method is evaluated with a computational study of 10 instances. The performance is compared with a full information (FI) method and another dynamic method (SI). The FI method evaluates the same instances, with all information known beforehand, thereby acting as a lower bound. The newly developed solution method has shown to perform 39% worse compared to full information, while being able to deal with dynamic requests. The SI method shows an increase of 60% in routing costs when compared with FI.
Next the development of the model and solution method, the implemented method is used to investigate the collaborative gain for four levels of collaboration in a dynamic environment. The level of collaboration is expressed by the number of requests that is reassigned between carriers. The results show that up to 23% collaborative gain can be realized in the dynamic setting.