Enrico Zio
Politecnico di Milano, Italy
Presentation Title: Risk for the Future, and the Future of Risk Assessment and Management
Abstract: Risk assessment and management is a mature scientific discipline, whose objectives are: identifying the hazards/threats which the system of interest is exposed to; analyzing potential accident scenarios, their causes and consequences; describing risk, possibly quantitatively and with an adequate treatment of the uncertainties associated to the assessment; and using the outcomes of the risk assessment for taking management decisions on preventive and mitigative measures against accidents and their consequences. The risk assessment, in particular, is based on the knowledge available on the system of interest.
On the other hand, our world is a technological one and in continuous transformation to meet the objectives and needs of efficiency, flexibility, sustainability etc, under significant social and environmental pressures. Innovations in technology are continuously being developed for the well being and to the benefit of all. These innovations generate systems and systems of systems, whose structural, logic and operational complexity continue to increase. In such evolving technological context, new and unknown hazards and threats emerge, which must be assessed to take appropriate decisions on system licensing, construction, operation, and on asset maintenance and management with the aim to prevent the occurrence of accidents and prepare to mitigate and recover from their consequences, were such accidents to occur.
In this lecture, I attempt to provide a partial view of the evolution of risk and to offer some directions of research and development of risk assessment and management, including:
- The use of simulation for effective accident scenario identification and exploration within the framework of computational risk assessment
- The role of artificial intelligence and machine learning in risk assessment and management
- The combination of complex network theory and input-output inoperability modeling for the risk assessment of systems of systems vulnerable to extreme wheather scenarios under climate change scenarios
- The exploitation of monitoring data for the dynamic updating of risk assessment and condition-based risk assessment
- The use of natural language processing for risk analysis
- The contribution of digital twins to adaptive risk assessment and management
- The extension of the framework of risk assessment to resilience analysis