Accidents in industry, medical clinics, and transportation lead to loss of lives and property. The costs for such accidents are very high to society, industry and individuals. Recent examples of high profile accidents include the Fukushima disaster and several aircraft accidents. Accidents in hospitals cost as much as $29 Billion per year (WHO estimate) in additional hospitalization, disability, lost income, medical expenses and litigation.
Looking at different industries, safety levels are very different. For example, there is a one in a million chance of getting harmed while in an aircraft. In comparison, there is a one in 300 chance of a patient being harmed during health care.
Current best practices to increase product safety uses methods which were developed during the period between World War II and 1970. They are essentially based on heuristics around designing and operating systems that were a lot less integrated and complex than current systems, and contained a lot more human participation in the system. A few science-based approaches have began to emerge in recent years.
This paper presents the hypothesis that in addition to the benefits proven in traditional design practice, the design axioms can form a scientific basis for an effective method to analyze designs from a product safety perspective.
The main cause of the 2005 Buncefield overfill accident in the UK is analyzed using an axiomatic approach to safety. Using the rigor of axiomatic design, it is possible to identify the shortcomings in the design that caused the accident. The same approach is then applied to an alternative solution. The result of this analysis is that the new design should be much more reliable than the previous.
In the conclusion, it is shown that it is possible to develop a science base for product safety based on Axiomatic Design. It is shown that such a science based approach enables design practitioners to approach unfamiliar safety related situations in a systematic way to identify risks and propose alternative designs with a lower risk level.