Abstract

The paper addresses the problem of efficiently planning routes for multiple ground vehicles used in goods delivery services. Given popularity of today's e-commerce, particularly under the COVID-19 pandemic conditions, goods delivery services have been booming than ever, dominated by small-scaled (electric) bikes and promised by autonomous vehicles. However, finding optimal routing paths for multiple delivery vehicles operating simultaneously in order to minimize transportation cost is a fundamental but challenging problem. In this paper, it is first proposed to exploit the mixed integer programming paradigm to model the delivery routing optimization problem (DROP) for multiple simultaneously-operating vehicles given their energy constraints. The routing optimization problem is then solved by the multi-chromosome genetic algorithm, where the number of delivery vehicles can be optimized. The proposed approach was evaluated in a real-world experiment in which goods were expected to be delivered from a depot to 26 suburb locations in Canberra, Australia. The obtained results demonstrate effectiveness of the proposed algorithm.