Using Aggregated Industry Accident Data, Bayesian Belief Network, and Human Reliability Analysis to Better Predict Human Error Probability in Aviation
Open AccessHuman error in aviation is a significant contributor to accidents, incidents, and unscheduled aircraft disruptions. Human error is responsible as a major causal factor (70% - 80%) in most aviation accidents (Wiegmann & Shappell, 2003). Traditional Human Reliability Analysis (HRA) models are simplistic in their approach to predict human error probabilities. Traditional methods ignore the dependencies among causal factors and don’t provide a casual picture of human error (Groth & Mosleh, 2012). This praxis uses aggregated industry accident data, a Bayesian Belief Network (BBN), and the Human Factors Analysis and Classification Systems (HFACS) to come up with a better estimation of human error probability in aviation. The HFACS provides a taxonomy that maps industry accident data into a BBN. The BBN method overcomes the shortcomings of the traditional human error analysis (Groth & Mosleh, 2012) by establishing the causal factors that affect operating crew behaviors to be identified and the dependency relations to be modeled in a BBN diagram. Using aggregated industry accident data adopted from (Moura et al., 2016) and literature, quantitative dependency of the causal factors (nodes) is specified in conditional probability tables. The resulting human error probabilities which are the posterior probabilities are computed using Netica software. From the human error probability assessed by BBN inference and sensitivity analysis, the major influencing factors responsible for human errors can be isolated and useful conclusions can be drawn that provide support for human error management.
- All rights reserved
Notice to Authors
If you are the author of this work and you have any questions about the information on this page, please use the Contact form to get in touch with us.