ABSTRACT
Traditional industrial automation systems developed under centralized architectures are statically programmed with determined procedures to perform predefined tasks in structured environments. The major challenges for these legacy systems are that they are unable to automatically discover alternative solutions, flexibly coordinate reconfigurable modules and actively deploy corresponding functions, to quickly respond to frequent changes and intelligently adapt to evolving requirements in dynamic environments. This paper presents a two-layer architecture modelling framework, including the high-level cyber module designed as multi-agent computing model and the low-level physical module designed as agent-embedded IEC 61499 function block model, to enable real-time adaptation at the device level and run-time intelligence throughout the whole system. The design results in a new computing module for high-level multi-agent-based automation architectures and a new design pattern for low-level function block modelled control solutions. The design is demonstrated and evaluated through various tests on the multi-agent simulation model developed in NetLogo and the experimental testbed designed on the Jetson Nano and Raspberry Pi platforms. The result shows that the design is feasible with improved performances and expected capabilities to respond to major challenges in Industry 4.0.
Acknowledgments
This work was supported by the Natural Sciences and Engineering Research Council of Canada (NSERC) under Grant CDE 486462-15.
Disclosure statement
No potential conflict of interest was reported by the author(s).