Initial considerations of a multi-layered run time assurance approach to enable unpiloted aircraft

Increased autonomy promises many advantages in the aviation domain, especially for unmanned aerial systems and small aircraft. However, several critical challenges remain and must be solved before highly autonomous or “unpiloted” operation can be fully realized. Of these challenges, safety and safety assurance is of utmost importance. Run-time assurance (RTA) has been shown, both theoretically and experimentally, to be a very promising avenue upon which to assure safety of the myriad of functions required for effective flight. This paper develops concepts based around an RTA architecture composed of multiple safety monitors encompassing and assuring many flight functions. Implementation of the concepts described in this work was performed on a small-unmanned air vehicle (SUAV) as a part of the NASA/FAA Traveler project. Lessons learned in requirements generation, system design, and operations are presented along with flight test results. These results are applicable to not only SUAV, but also to small manned aircraft and other vehicle systems where safety assured autonomy is desired.