Objective: The process of understanding the causes of adverse events associated with complex engineered systems can be time consuming and expensive. It often requires substantial human and physical resources ranging from a few engineers up to multiple teams of domain specialists from collaborating organisations. The research presented in this article aims to provide more effective support to the analysts involved in root cause analysis (RCA) by exploring the combined application of the Issue Based Information System (IBIS) and the Function Analysis Diagram (FAD) methods. The first method (IBIS) introduces the concept of argument-based rationale for explicit justification of the nodes of a cause-effect chain as well as of redesign decisions, while the second method (FAD) introduces the notion of structure-dependent functional modelling of complex systems in normal and failure states.
Method: Causation data from publicly available technical reports of the Space Shuttle Challenger disaster was reverse-engineered using a root cause analysis approach based on the IBIS and FAD notations. IBIS and FAD were implemented using a free and open source software tool known as designVUE. The approach was evaluated by comparing it to a method for root cause analysis widely used in industry and assessing how it satisfies generic requirements for root cause analysis.
Results: The results show that the proposed IBIS-FAD approach provides a rich description of the causes for an accident presented in a manner that facilitates information access and understanding. The IBIS notation allowed for explicit modelling of the reasons supporting or refuting failure hypotheses along with evidence. The FAD notation provided a clear and concise method to visualise the complex set of non-linear interactions leading to the failure of a system by annotating graphical schematics of the design with the functions exchanged between its components. Finally, the results show that the approach supports the capture and justification of redesign decisions and ties them to initiating problems in a way that promotes the prevention of accident re-occurrence.
Conclusions: Argument-based rationale with IBIS and FAD-style functional modelling are powerful concepts to extend the tool set available to support the root cause analysis process. The approach proposed in this article provides a unique tool that would be of value to academics, practitioners, and regulators concerned with root cause analysis and opportunities to improve the process of understanding adverse events.