Semi-Automated Event Specification for Knowledge-Based Event Detection

Abstract

Cyber-Physical Systems represent a fusion of spatially distributed digital and physical elements. Due to the increased complexity of such systems, understanding their holistic operational conditions becomes more and more challenging, especially for non-skilled operators. Event detection is a proven technique in this problem domain. However, manual event specification can be laborious and error-prone, motivating automated approaches. This paper proposes a knowledge-based event detection system that leverages abstract reasoning rules to derive time series event specifications from a system model. By automating the derivation of specifications, system integrators can focus on crafting event specifications for the system's intricate parts. Leveraging Semantic Web technology and Signal Temporal Logic, we evaluate the feasibility of the proposed approach in an electric vehicle charging scenario.