Ontology-Based Holonic Event-Driven Architecture for Autonomous Networked Manufacturing Systems
Hui Cao,Xing Yang,Raoyi Deng
DOI: https://doi.org/10.1109/tase.2020.3025784
IF: 6.636
2021-01-01
IEEE Transactions on Automation Science and Engineering
Abstract:The current trends in personalized products drive the paradigms of production systems toward autonomous networked manufacturing systems. This article proposes an ontology-based holonic event-driven architecture for implementing loosely coupled, holonic, autonomous distributed systems. The event-driven architecture (EDA) enables the services provided by different organizations and their suborganizations to be autonomously configured and integrated, while ensuring that the predefined organizational event access rules are compulsorily followed. This is realized by organizing the event services in a holonic manager and modeling the knowledge of services via the ontology developed in this study. The ontology model and the corresponding autonomous configuration mechanism are introduced. We developed a demonstration case to elaborate our approach and implemented a testbed using Java-based technology for testing the effectiveness, efficiency, scalability, and reliability of the proposed approach. Note to Practitioners-The ever-changing customer demands, the volatile market, and the trend of mass personalization require manufacturing systems to be smart in the event of uncertainties. To this end, information sharing in the supply chain is a key factor. However, there are several barriers in implementing a cross-organizational information system, including security, access privileges, and system interoperability. To lift these barriers, this article proposes an ontology model and a hierarchical EDA. Compared with conventional architectures, our approach facilitates the management of access privileges, integration of interorganizational information systems, and their interoperation. Moreover, our approach is competitive compared with the conventional service-oriented architecture in terms of both scalability and reliability of the system. Before the proposed approach can be practically applied, the ontology model needs to be further extended, and the event registration and message exchange protocol need to be standardized.