Design and Testing of an Approach to Automated In-Flight Safety Risk Management for sUAS Operations
Ersin Ancel, Steven D. Young, Cuong C. Quach, Rafia F. Haq, Kaveh Darafsheh, Kyle M. Smalling, Sixto L. Vazquez, Evan T. Dill, Ryan C. Condotta, Bailey E. Ethridge, Logan R. Teska, Thomas A. Johnson
Abstract
View Video Presentation: https://doi.org/10.2514/6.2022-3459.vid An onboard risk management automation design is presented based on run-time assurance principles, as well as the concept for In-Time Aviation Safety Management Systems (IASMS) as described by the National Academies. The automation is designed to operate independently of the autopilot and perform real-time risk assessment spanning multiple classes of hazards, predict constraint violations, and track autopilot states. In the event of elevated risk conditions or predicted constraint violations, the automation will select from a set of available contingencies and trigger autopilot mode changes if necessary to mitigate risk exposure. The onboard automation also informs the remote operator/pilot of what the independent monitor is observing and any contingency decisions or actions that may arise during flight. Details of an implementation of this design and results of verification and validation activities, as required to meet stringent NASA software and system assurance standards, are also presented. This includes simulation and flight testing using small unmanned aircraft systems.