A Bayesian approach to system safety assessment and compliance assessment for Unmanned Aircraft Systems

This paper presents a new approach to showing compliance to system safety requirements for aviation systems. The aim is to improve the objectivity, transparency, and rationality of compliance findings in those cases where there is uncertainty in the assessments of the system. A Bayesian approach is adopted that facilitates a more comprehensive treatment of the uncertainties inherent to all system safety assessments. The assessment and compliance framework is reformulated as a problem of decision making under uncertainty, and a normative decision approach is used to illustrate the approach. A case study system safety assessment of a civil unmanned aircraft system is used to exemplify the proposed approach. The proposed approach could be readily applied to any regulatory compliance process and would represent a significant change to, and advancement over, current aviation safety regulatory practice. This paper is the first to describe the application of Bayesian techniques to the field of aviation system safety analysis. The adoption of the proposed compliance approach would bring aviation system safety practitioners in line with more contemporary (and well established) approaches adopted in the nuclear power and space launch industries.