An empirically benchmarked human reliability analysis of general aviation. (February 2020)

• Record Type: Journal Article
• Authors: Burns, Kevin; Bonaceto, Craig
• Is Part Of: Reliability engineering & system safety. Volume 194(2020)
• Publication Date: 2020-02
• Subjects: Accident rate data -- General aviation -- Human error probability -- Human reliability analysis -- Performance shaping factors -- Probabilistic risk assessment; Reliability (Engineering) -- Periodicals; System safety -- Periodicals; Industrial safety -- Periodicals; Fiabilité -- Périodiques; Sécurité des systèmes -- Périodiques; Sécurité du travail -- Périodiques; 620.00452
• DOI: 10.1016/j.ress.2018.07.028
• Languages: English
• ISSNs: 0951-8320

Abstract

Highlights: General aviation risk model matches historical accident rate data to within 5%. Nominal pilot error probabilities range from 2.5E-07 to 3.0E-04. A Digital Copilot reduces the accident rate by 20%. Abstract: A Human Reliability Analysis of general aviation is empirically benchmarked, using Probabilistic Risk Assessment methods and historical accident rate data. The analysis posits three types of pilot actions, namely Knowledge-based, Rule-based, and Skill-based, each with a nominal human error probability treated as an unknown parameter. Various performance shaping factors are treated as known multipliers to these probabilities in quantifying pilot-related accident sequence equations. The equations are aligned with accident frequencies reported in a general aviation safety database, which includes nearly one thousand accidents in over twenty million flights annually. These equations are then solved for values of the three variables, thereby establishing an empirical basis for the assumed types of actions and associated performance shaping factors. The resulting model, benchmarked against retrospective data, is also used in prospective fashion to quantify the ergonomic impacts and safety benefits of a prototype system that provides pilots with cognitive assistance in general aviation.