Machine Learning in Cyber-Physical Systems in Industry 4.0

摘要

Cyber-physical systems (CPS) have emerged with development of most great applications in the modern world due to their ability to integrate computation, networking, and physical process. CPS and ML applications are widely used in Industry 4.0, military, robotics, and physical security. Development of ML techniques in CPS is strongly linked according to the definition of CPS that states CPS is the mechanism of monitoring and controlling processes using computer-based algorithms. Optimizations adopted with ML in CPS include domain adaptation and fine tuning of current systems, boosting, introducing more safety and robustness by detection and reduction of vulnerabilities, and reducing computation time in time-critical systems. Generally, ML helps CPS to learn and adapt using intelligent models that are generated from training of large-scale data after processing and analysis.