A report on a Project Submitted in Partial Fulfilment of the Requirements for the Postgraduate Diploma in Safety Critical Systems Engineering. Presented to the Department of Computer Science of the University of York by James Inge.
The behaviour of many complex systems is based not only on their static design but also on configurable data used by the system. In order to assess the safety of such a system, it is necessary to have an understanding about the types of problem such use of data may cause, through a process of safety analysis.
This report reviews current guidance and best practice for treatment of data in safety analysis. It finds that while advice exists on how to manage the safety impact of data use after a potential hazard has been identified, there is a lack of tools and guidance for the initial task of hazard identification for safety-related data.
A taxonomy of types of data fault is proposed that can be used as a checklist to aid in the hazard identification process. The taxonomy is then validated using accident investigation reports, to determine whether it is capable of classifying the data-related issues seen in real life.