Agent Based Modelling for Migration of Industrial Control Systems

• Published in: 2019 IEEE International Conference on Systems, Man and Cybernetics (SMC) pp. 3292 - 3297
• Main Authors: Mohan, Devina, Chandrika, K.R., Sudarsan, Sithu D.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01-10-2019
• Subjects: Aging; Computer architecture; Computer languages; Control systems; Integrated circuits; Servers
• ISSN: 2577-1655
• DOI: 10.1109/SMC.2019.8913893

Keeping industrial control systems on par with state-of-the-art is required to (i) make industrial plants greener with reduced energy consumption and emissions, (ii) improve productivity due to better optimization and automation, and (iii) reduce maintenance cost due to improved hardware and software. However, updating and upgrading Industrial Control Systems is complex, time consuming and resource intensive task because of execution order constraints and hard real time needs. Added challenge in migrating control logic from legacy and heritage systems is shortage of Subject Matter Experts with prior knowledge of such heritage and legacy systems. This paper proposes a migration system based on Agent Based Modelling (ABM) architecture to map the control entities and parameters from source to target system enabling easier migration of legacy systems to newer versions. It provides a standard platform where knowledge of the experts is captured and stored for re-use, improving confidence level of suggestions, training generic automation engineers on system specific aspects as well as support handling unknown source systems. The proposed migration system is flexible, scalable and usable because of ABM architecture. It is capable of creating virtual agents from micro-agents offering atomic and/or granular functionality.