Abstract

The EU project InfraStress aims to provide solutions and support for the Sensitive Industrial Plants and Sites (SIPS) and Critical Infrastructure (CI) in related to a number of physical and cyber threats, including technological hazards and extreme natural hazards (NaTech). The SIPS usually consist of large facilities, notably processing chemical plants where large amounts of dangerous substances are present. The InfraStress approach to risk modelling that also considers model updating with data from the specific sensors on hazards/threats is presented in the paper. First, the main steps in the risk modelling are explained, including risk communication using Attack Tree notation, as well as use of the Bayesian Belief Networks (BBNs) to elaborate a quantitative probabilistic risk model. Next, the use of the Kafka broker system among the InfraStress set of components is presented, including some details of data exchange with the BBN software API. In the illustrative example, we present the case procedure and results following the InfraStress risk modelling approach, where we (via data update in the model nodes) consider updated ("live") info from the sensors. In our understanding, the proposed InfraStress risk modelling using live data could provide valuable aid to the SIPS risk managers in evaluating the information from the sensors and its analysis and interpretation in the underlying risk models.