Agent-based negotiation and decision-making for efficient hinterland transport plan

The logistics and transport operations, which have become more information intensive and more technologically dependent, are undergoing dramatic revolutions. Among various logistics activities, the role of transport is becoming more and more important. In order to provide high quality service and obtain higher return on investment, information sharing and cooperative decision making become a must in this dynamic changing environment. In particular, multi agent (MA) technologies have drawn significant attention for enabling autonomous information sharing and cooperation in a complex and distributed environment. This paper presents the development of coordination and decision making mechanisms for the implementation of autonomous control with MA technology in the domain of hinterland transport. It tackles the challenges of the hinterland transport planning caused by limited information sharing and lack of cooperation. The objective of the developed system is to provide quality transport plan to achieve high level of performance and robustness in hinterland logistics. The paper specifies and implements the MA cooperation, negotiation and decision-making processes. A case study is presented to illustrate the effectiveness of the developed system.

I. Introduction

Among various logistics activities, the role of transport is becoming more and more important. The efficiency and effectiveness of the transport plan directly link to the logistics service satisfaction. Automating transport plan process(TPP) is remain a challenging task because the logistics network tends to be increasingly complex, dynamic and distributed. Conventional approach to TPP is optimization oriented by means of centralized control. It tries to optimize certain objectives using the methods from the field of operational research which includes VRP, TSP, flow job shop schedule and etc. The obtained solution primarily focuses on individual objectives. In real logistics transport network, however, different entities (such as physical resources, human resources, business participants) are economically interconnected and physically distributed. Central coordination is thus no longer feasible for automate planning and controlling of a complete logistics system. Confronted with the challenges, the concept of autonomous logistics [1] has been put forward. It utilizes the MA technology to facilitate the information exchange and decision making process and leads to a more efficient solution.