An effective real time GRASP-based metaheuristic: Application to order consolidation and dynamic selection of transshipment points for time-critical freight logistics

Abstract

Time-critical freight logistics is an area within logistics research where the shipper’s orders need to be received relatively urgently using a third party logistics (3PL) that provides a quote (bid) to the shipper within a very short period. We solved this 3PL problem by developing an effective meta-heuristic based on the Greedy Randomized Adaptive Search Procedure (GRASP). This is achieved by introducing novel attributes in the construction of the restricted candidate list while incorporating flexible and intelligent rules, some of which are inspired by expert knowledge. The approach performs order consolidation, locates transshipment points and performs an optimal assignment of shipments to the selected consolidation points dynamically and in real time. This intelligent system embeds expert knowledge within the design of neighbourhood reduction schemes and data structures to speed up the search. This is achieved by recording computed data that does not need to be recomputed again while avoiding unnecessary computations of the non-promising alternatives.The performance of this real time optimisation and scheduling tool is tested with a European 3PL company over a 13 weeks period in late 2017 resulting in a significant cost saving and a considerable reduction in CO2 emissions. This powerful decision support system assists the 3PL company in gaining competitive leadership advantage through producing promising quotes that turn customer requests into real customer orders.