Assuring Autonomy Ethics, Trust, and Safety in DAF Mission Critical Systems with Air Force Research Labs and the Lifecycle Management Center

Rapidly developing commercial machine learning (ML) and artificial intelligence (AI) technologies offer significant capabilities for the U.S. DoD and specifically the U.S. Air Force. The DoD has identified that incorporating ML into both legacy and future defense systems is critical for the U.S. to maintain overmatch among near-peer and peer adversaries including Russia and, most prominently, China. The ability to operate in fast-paced, contested, nonpermissive, force-on-force engagements, particularly under degraded conditions, drives the need for increased autonomy. This, in turn, relies on ML technology for increasingly complex decisions, mission-critical actions, and maintaining trustworthiness and safety. The Air Force and Department of Defense have a critical need for utilizing AI/ML safety assessment and evaluation testing software for unmanned flight-testing systems, particularly for aircraft such as the XQ-58 and other developing autonomous aerial systems. With the growing reliance on unmanned autonomous systems in military operations, it is crucial to ensure the safety of flight and mitigate potential in-flight issues that may arise from autonomous systems. The Department of the Air Force requires a reliable and effective safety assessment evaluation software to ensure that autonomous flying aircraft operate safely and with utmost trust. The nLoop software is capable of assessing the safety and reliability of autonomous vehicles and their software in various edge cases, including scenarios that are not typically encountered during normal operation or testing.