Knowledge elicitation through collaborative modelling: A case study of the British railway industry

A vast range of data exists within the railway industry, and their availability continues to increase as a result of uninterrupted data collection processes across the industry. A recent research project has demonstrated the feasibility of using the data available within the railway industry to inform new mechanisms to assure safety and security of customers, staff and the public in an industry where the interdependence between physical and digital environments is set to grow exponentially over the next few years. The project delivered a small-scale solution which serves as a proof of concept for an approach to understanding the factors potentially affecting safety of complex rail systems. Elicitation of experts' knowledge of railway data and their importance for safety was therefore the key to its success. A collaborative modelling approach to knowledge elicitation from experts was used in this project. The approach, which had been applied in organisations in the past, was then applied to an industrial scale problem, where both the ownership of both the problem and the related knowledge are distributed and disparate. In this context the approach overcame many of the known challenges of knowledge elicitation and successfully produced, among other outcomes, a series of models of the knowledge domain. This paper describes the mechanism used to motivate and guide a group of key data and safety management experts from the British railway industry to contribute with their expertise to the development of a solution to understanding safety in tomorrow's railway.