HazOp Scenarios Modelling for Digital Factories

The Hazard Operability (HazOp) method aims to identify the potential hazards present in a plant in order to ensure its safety and prevent damage. The process involves identifying the causes of these hazards and their potential consequences and designing effective measures to minimise the probability of their occurrence. Inductive Logic Programming (ILP) is a technique used for supervised learning, where the aim is to construct a theory that explains a given set of observations. In [1], we proposed a learning technique based on ILP. It enables general diagnostic rules to be established to guide the operator in managing future malfunctions. In this paper, we carry on our work on Onto-Relational Learning by developing the SteamBoiler domain ontology using the RAMI top-level ontology and the HAZOP method. RAMI serves as a foundation for Industrie 4.0, offering a standardised interpretation of terminology. Further development and adaptation of each domain are executed at lower levels. The ontology developed is considered prior conceptual knowledge in ILP to induce supervision rules.