Fair Tour Planning : Managing Parcel Logistics by Neural Networks

In times of high-frequency eCommerce, the value of parcel logistics grows continuously. Customers expect “sofortness” (instant delivery), which increases pressure on the whole delivery system. The ultimate performance driver is workforce. Teamwork is essential because only one conveyor belt exists to handle the whole process from collection to delivery; the “last mile” is the most complex part of the logistics process that must support fairness, rules, and compliance. We developed a platform to fulfill these requirements. An artificial neural network with deep learning characteristics builds clusters of logic and geographic segments. It helps to plan the delivery and to calculate the necessary performance per parcel and delivery stops. Daily controlling, automatic data preparation, and enrichment feed the artificial neural network. Finally, the application places segments sequentially to meet team fairness in a highly agile daily business of a logistics company in Luxemburg.