Hugin Expert

This project investigates means of moving from passive decision support systems to active solutions which can deal with the complexity inherent in decision making processes. The project comprises six partners and a user group.

The project addresses the safety justification process for software-intensive systems. The primary aim of the work is to develop a means of using Bayesian Belief Networks (BBNs) to reason about the types of information contained in a safety argument. Existing methods such as fault tree analysis, which work well for hardware systems, do not fully support the evaluation of evidence about the nature of the design process, as required for software intensive systems. The use of BBN technology will allow qualitative reasoning about design errors and linking of conclusions to the frequency-based data demanded by fault-tree analyses. The objectives of the project are to develop a method for constructing software safety arguments using BBNs, to adapt an existing BBN tool to support the method and to evaluate the application of the method and tool through practical trials. The results are the provision of a manual detailing procedures for identifying and structuring evidence that a system meets the safety requirements of IEC 1508, the provision of a tool using BBN technology to automate the implementation of the method, and a quantified comparison of the performance of the proposed method and tool compared to conventional methods. A practical approach is adopted, based on modifications to an existing BBN tool, together with the involvement of experienced practitioners to formulate strategies for representing safety arguments. Practical trials of the method and tool will be carried out; the involvement of a user group throughout the work programme is an integral part of the project - which is intended to achieve good exploitation of the work. The potential impact of the project for the European Community in terms of costs, time-to-market and safety standards is high. European legislation requires industry to justify that its products are safe. Currently, there is little experience of the necessary techniques within industry and significant costs can be accrued in developing safety cases, potentially excluding small companies from the marketplace. The availability of practical methods and tools should assist companies in reaching the required European safety standards in a cost effective manner. The strategy for exploitation is through the involvement of end users throughout the project. The user group and three of the partners are involved in the construction of safety arguments as operators of safety related systems or as independent assessment organisations. To reduce technical risks, the products of the project can be exploited at three levels as follows: 1. Provision of guidance on construction of a safety argument; this is needed by industry and can be developed regardless of the specific techniques used, hence risks of poor exploitation are low. 2. Provision of procedures for providing a safety case based on BBN technology; this can be developed regardless of whether the automated tool support is available and hence exploitation risks are seen as low. 3. Provision of the automated tool support; successful exploitation at this level will achieve the greatest benefit in terms of reducing industry's costs, but is also at greatest risk of weaknesses in the tool support; nevertheless the tool is well proven and hence the risks to exploitation are seen as moderate.