An Ontology-Based Approach for Risk Evaluation in Human-Machine Interaction

Abstract

The Occupational Safety and Health areas have been changing to address the human-machine interaction (HMI) in collaborative workplaces. As regard, the Preliminary Hazard Analysis sub-area must identify, monitor, and evaluate new risks which emerge from such interaction. Although the classical HAZOP method covers this matter, there are many ambiguities in the hazardous scenario descriptions. Thus, cyber-physical systems come into play to help using computer technologies including ontologies, but the approaches do not have an expressive ontology capable of evaluating the risks covering the hazardous and HMI contexts. Therefore, this work aims to propose an application ontology to evaluate risk situations in a collaborative environment, capable of describing all the entities and their relationships. The proposed ontology was evaluated in a simulated scenario, being able to model different hazardous situations in HMI collaborative environments, and supporting the detection and characterization of hazardous events based on the distance between machines and operators.