Interpreting chain of events for safety analysis

Because of difficulties of safety analysis in large systems and the complexities of managing large amount of data in these systems, the need for a supporting system has become an important area of research. Managing requirements, components and failure modes of a large system for safety analysis without tool support is difficult and could result in skipping or missing details which may cause an accident.The main goal of this project is to develop a system model for safety analysis facilitation. Sequences of events in a system and its environment may cause an accident in the system?s environment. People often have problem following long cause-consequence sequences of events, while accidents with a short path from initiating event to accident are easy to identify. The system model developed in this project enables automatic generation of event sequences that can cause an accident to the environment of a system. For this purpose, the system model uses the domain ontology as its knowledge base. This ontology must contain cause-consequences reported from a safety expert. A rule engine is used for reasoning about these cause-consequence concepts and generating event chains. The process of developing this system model, how the system model uses a domain ontology as its knowledge base and how the cause-consequence concepts should be added to the domain ontology are described in this project. Finally, the developed system model is tested with a real world example (a simplified steam boiler) and the expected event chains generated automatically.