Rolling horizon planning for a dynamic collaborative routing problem with full-truckload pickup and delivery requests

Abstract

In order to improve their operational efficiency, small and mid-sized freight forwarders can establish horizontal coalitions in order to exchange customer requests. Decentralized operational transportation planning processes enabling request exchange among forwarders in spite of information asymmetry and distributed decision-making competences is referred to as collaborative transportation planning (CTP). CTP can help forwarders take advantage of economies of scale and reduce their costs of fulfilling customer requests compared to the case of isolated planning without request exchange. In order to exploit the potential of cost-savings embedded in CTP, appropriate request exchange mechanisms have to be developed. In this paper, the dynamic CTP problem of a coalition of freight forwarders serving full-truckload transport requests is studied. Two rolling horizon planning approaches are proposed to solve the dynamic routing problems. It is analyzed how the planning results, in particular the cost reduction realized by CTP, are influenced by different planning settings. Computational experiments show that the planning results of CTP are considerably superior to those obtained by isolated planning, and the realized cost-savings in percentage remain relatively constant, independently of the test settings.