Abstract

Events recognition is central to occupational health and safety OHS, since the system can selectively start proper prediction services according to the user current situation and past knowledge taken from huge databases. In this sense, a fusion framework that combines data from multiples sources to achieve more specific inferences is needed. Our proposed model provides the big picture about risk analysis for that employee at that place in that moment in a real world environment. Our main contribution lies in building a causality model for accidents investigation by means of a well-defined spatiotemporal constraints on offshore oil industry domain.