Dynamic scheduling in RFID-driven discrete manufacturing system by using multi-layer network metrics as heuristic information

Jiewu Leng,Pingyu Jiang

DOI: https://doi.org/10.1007/s10845-017-1301-y

IF: 8.3

2017-02-07

Journal of Intelligent Manufacturing

Abstract:Abstract Since discrete manufacturing system (DMS) is a complicated dynamic network that comprises of processes, machines, and work in process, a coherent methodology for performance tracking and sustainable improvement at the system/network level is of great significance for manufacturers to respond rapidly in a mass customization paradigm. Fortunately, the radio frequency identification (RFID) technologies provide us the real-time tracking ability of the production process that suffers unpredictable and recessive disturbances. This paper proposes a dynamic scheduling approach based on multi-layer network metrics of RFID-driven DMS. Firstly, considering the elements of DMS (e.g., parts, manufacturing activities, and equipment) and relationships among them, a DMS model named complex manufacturing network (CMN) is proposed. Then, several multi-layer network metrics of the CMN are defined and analysed. The implications of these metrics lead to a better understanding of the current status and performance of DMS. Thirdly, a dynamic scheduling algorithm using these metrics as heuristic information is proposed to solve multi-resources and independent-task DMS. Finally, a Printing Machine manufacturing system is chosen as an example to illustrate the feasibility of the proposed approach.

engineering, manufacturing,computer science, artificial intelligence