Modelling and Performance Evaluation of Containerised Parcel Delivery

Abstract

This paper investigates different factors that affect the performance of containerised transportation in parcel delivery networks. Motivated by a situation facing a postal delivery company in Australia, we study how container utilisation rate, sortation activities, and changes in cost parameters can affect the overall performance of a parcel delivery network. Mixed-integer programming and machine learning are employed to model a realistic parcel delivery network considering sortation activities and to evaluate the performance of this network using data from a major postal service provider. The findings of this study can help parcel delivery companies to make more informed investment decisions and introduce more effective performance improvement initiatives.