Toward monitoring organizational safety indicators by integrating probabilistic risk assessment, socio-technical systems theory, and big data analytics

Many catastrophic accidents have organizational factors as key contributors; however, current generations of Probabilistic Risk Assessment (PRA) do not include a comprehensive representation of the underlying organizational failure mechanisms. This paper reports on the current status of new research with the idealistic goal of monitoring organizational safety indicators. Because of the evolving nature of computational power and information-sharing technologies, ‘Internet of Things’ has been adopted as a metaphor to describe the authors’ vision for combining multiple levels of organizational analysis into a real-time application for monitoring the changing landscape of risk. The short-term objectives are: (1) identifying the organizational root causes of failure and their paths of influence on technical system performance, utilizing theoretical models of social phenomena, (2) quantifying the models using advanced measurement and predictive methodologies, big data analytics, and uncertainty analysis, and (3) proposing preventive approaches. Socio-Technical risk analysis deals with wide-ranging, incomplete, and unstructured data. Therefore, this research focuses on developing hybrid predictive technologies for PRA that are not only grounded on Socio-Technical Systems theory, but also serve to expand the classical approach of data extraction and execution for risk analysis by incorporating techniques such as text mining, network data analytics, and data curation.