Author:

Keywords:

HLES, cross-docking, schedule execution

Abstract:

The coordination and control of cross-docking operations is a complex task and a challenging problem. A severe competition in the logistics sector and ever-increasing traffic (congestion) make that cross-docks should be able to operate efficiently in uncertain and dynamic environments. Moreover, the operational coordination and control is a going concern, so 'one-shot optimization' is not sufficient.As the logistics and manufacturing domain are characterized by similar properties (e.g. large decision space, nonlinearity, uncertainty, etc.), this thesis applies the concepts and principles of the Holonic Manufacturing Execution System (HMES) in order to develop a Holonic Logistics Execution System (HLES). The HLES is in charge of the organization of the logistic operations (e.g. resource allocation, tracking and tracing, etc.). By combining the PROSA reference architecture and the delegate MAS pattern in its software architecture, the HLES is able to provide a view on the expected short-term future of the system. The provided visibility is a valuable property for the staff, and also allows the control system to take better (informed) decisions.The objective of this thesis is to show that the HLES concept is a valuable option for a cross-docking logistics execution system. To this end, the thesis explains how such an HLES implementation can be developed. First, several adaptations and extensions to the latest HMES implementation are proposed. The HLES consists of a reusable core, domain-specific models and application-specific decision mechanisms. The introduced adaptations to the core allow the HLES to offer support for mobile resources (e.g. trucks), batching and multi-resource allocation. The thesis also presents the necessary models and decision mechanisms for the entities relevant in the context of cross-docking. It has been shown by simulation experiments that the holonic technology is applicable and that the developed system works as intended.Second, to support the cross-docking HLES in finding good global solutions, a vehicle routing scheduling system and truck scheduling system are developed. A cooperation mechanism between the holonic on-line control system and these scheduling systems is also presented. As the developed scheduling systems make use of a simplified and approximated model of reality, the schedules provided by these systems can be (partially) infeasible. The HLES is however able to compensate for several simplifications. Moreover, the HLES also deals with deviations from the schedule, making the scheduling approach more robust. Simulation experiments confirm that the cooperation between the HLES and external scheduling systems improves the performance.