Decision analytics for safety risk modeling of unmanned aircraft systems (UAS)

Manufacturers and operators of Unmanned Aircraft Systems (UAS) in the United States are keenly interested in gaining access to the National Airspace System (NAS) for a variety of emerging multi-faceted missions, such as border patrol, crop monitoring, firefighting, pipeline surveillance, scientific experiments, etc. However, since these air vehicle operations are relatively new compared to manned aviation, non-military accident and incident data are extremely rare, so alternative modeling approaches to conventional fault tree and event tree logic are required to understand the impact of the introducing these operations into the NAS. While alternative real-time and fast-time simulation modeling research for UAS in the NAS is a major focus area of the aerospace industry and government agencies, the development of complementary probabilistic analytical methods and tools needs to similarly advance. Decision analytics are presented to integrate a nascent hazard taxonomy with probabilistic risk modeling. The safety risk modeling approach initiates with a hypothetical scenario where the hazardous elements from the UAS, Operations, Airmen and the Environment are identified. Subsequent modeling steps involve the creation of a Bayesian Belief Network (BBN) that includes proposed mitigations and sensitivity analyses. The model is demonstrated with a UAS pipeline monitoring scenario.