Aerospace Engineering Seminar Series: Karen Feigh "Towards Safe Collaboration Between Autonomous Pilots and Human Crews for Intelligence, Surveillance, and Reconnaissance & Medivac"

Speaker Bio: Dr. Karen M. Feigh is the Daniel S. Lewis Professor and Associate Chair for Research at Georgia Tech’s Daniel Guggenheim School of Aerospace Engineering with a courtesy appointment in the School of Interactive Computing. Dr. Feigh’s research integrates aerospace engineering, cognitive systems engineering, and human factors to improve human-machine collaboration in complex, safety-critical domains such as aviation, spaceflight, and autonomous systems. Her work addresses shared mental models, judgment, and situation awareness in human-AI teams; inferring decision strategies from behavioral data; and modeling trust in human-machine interactions. Recent projects include safe collaboration between autonomous pilots and human crews, the role of shared mental models in human-AI teaming, and methods for inferring belief states in partially observable human-robot teams. She has served as an Associate Editor for the Journal of Cognitive Engineering and Decision Making and the IEEE Transactions on Human-Machine Systems, helping to shape the research agenda in human-automation interaction. Dr. Feigh has also contributed her expertise to the National Academies of Sciences, Engineering, and Medicine, serving on committees addressing aviation safety, human-autonomy integration, and the future workforce for aerospace systems. Through her interdisciplinary research, leadership, and mentorship, Dr. Feigh advances the design of resilient socio-technical systems and prepares the next generation of aerospace engineers to address the challenges of increasingly complex human-machine partnerships.

Abstract: Many aviation missions today are accomplished by a heterogenous crew of pilots and mission specialists. As fully Automated Pilots (AP) are integrated into aviation crews, effective teaming will be necessary for safety assurance and mission effectiveness. The talk will present a series of flight simulator studies that explored teaming between a non-pilot human operator and an AP collaborating on a maritime Intelligence, Surveillance, and Reconnaissance (ISR) and Medivac missions. The ISR studies compared a baseline Waypoint AP behavior, requiring human intervention in aircraft control to prevent overflight of damage-causing enemy ships, with collision avoidance autonomy using control barrier functions. The Medivac missions compared the impact of visual v. auditory modalities. In both mission types, performance, workload, situation awareness and trust are measured.