Machine Learning Driven Digital Twin for Industrial Control Black Box System: A Novel Framework and Case Study

Industrial control systems are excessively used in advanced manufacturing environments. The lack of information and data regarding the internal workings of certain systems makes virtual modelling for their Digital Twin challenging. As a result, these systems are often classified as “black box“ systems. There is minimal research found on DT models for industrial control black box systems. Therefore, a novel algorithm to model the Digital Twin of the industrial control black box system in the cyber domain has been presented in this paper. Machine Learning techniques were used to develop a high-fidelity Digital Twin model of a black box system. Real-time sensor data were recorded and used to validate the proposed novel algorithm. This paper presents the proposed algorithm's effectiveness in developing a robust Digital Twin model of industrial control back box system.