An international industrial group entrusted Proxem with all incident logs recorded by employees in these various plants. Semantic analysis of these logs entailed detecting, for each incident, the problems found, the factors identified, and the likely causes of the incident. The elements detected could then be quantified and compared, making it possible to determine whether their frequency increased or decreased over time. Semantic analysis was the basis for establishing a global view of a company’s production plants.
Data visualization helped make sense of large amounts of incident logs that were otherwise unintelligible. In particular, it highlighted causal chains between incidents, as well as key factors, materials and equipment at risk, and less secure production centers. It also made it possible to see incidents that occurred with an abnormal frequency and incidents that occurred simultaneously or sequentially and were highly correlated