A Workflow for Designing an On-line Supervisor for Cyber-Physical Systems: a Case Study

Cyber-physical systems are expected to play a central role in future smart computing solutions, as they will fill the gap with the physical world. In this context, the safety of cyber-physical systems is critical to ensure a smart system resilient from malfunctions or unpredictable events. In this paper we present a workflow to design an on-line supervisor that aims at enforcing safety rules on a cyber-physical system. The workflow is based on the digital-twins concept: the model of the system is simulated to derive the safety rules that characterize its correct behavior. Such safety rules are then exploited by an online supervisor service that collects and analyzes state data from the cyber-physical system to detect if the system is currently in violation of the safety rules. In case violations are detected, the service sends commands to the system in order to mitigate its effects. In order to demonstrate the feasibility of the proposed approach, a case study of a simple line follower robot is presented.