A civil aviation safety assessment model using a Bayesian belief network (BBN)

• DOI: 10.1017/s0001924000000907
• Publication Date: 2016-02-05T09:04:26Z
• Authors: R. Greenberg; S. C. Cook; D. Harris
• Keywords: 0502 economics and business; 05 social sciences; 0202 electrical engineering, electronic engineering, information engineering; 02 engineering and technology

AbstractIn this paper we present a Bayesian belief network (BBN) socio-technical model for investigating the accident rate for multi-crew civil airline aircraft. The model emphasises the influence of airline policy and societal behaviour patterns on the pilots within the piloting system. The main claim of this paper is that a BBN can be used to bring most aviation safety-critical elements into a common quantitative safety assessment despite the unique problems posed by the very low probability of accidents. We support this claim by replicating certain phenomena such as the low accident rate, the difference between the ‘more’ and ‘less’ safe airlines and other statistical factors of civil aviation. In particular, the model succeeds in explaining the large gap of six to seven orders of magnitude between in-flight measurements of pilots’ error and accident rate.