Recurrent learning for drones improves safety after remote control interruptions
Interruption-Aware Recurrent Reinforcement Learning for UAS Navigation Under Temporary Overrides
Artificial IntelligenceRobotics
Summary
Drones often use smart control systems to fly safely and avoid crashes, but when a human pilot or an automatic system temporarily takes over control, the drone’s learned memory can get confused. The authors found that after these interruptions, drones are more likely to get into risky situations. To fix this, they designed new smart control systems that remember interruptions explicitly and adjust better when control is returned, making drone flights safer after handovers. Their tests show these methods cut collision risks and mission failures significantly without making normal flying worse.
What this means in practice
- •For drone system engineers: Build drone control software that handles temporary pilot or system takeovers safely and reduces risks after switching back.
- •For autonomous vehicle developers: Design recurrent learning policies for self-driving systems that remain robust when interrupted by human or safety overrides.
Authors
Christopher Conrad, Deepak Panda, Weisi Guo, Antonios Tsourdos
Abstract
Safe autonomous navigation of UASs requires tactical deconfliction to address conflicts that strategic coordination cannot prevent. This task is challenging under partial observability, when sensor or traffic data are missing, delayed, or degraded. Geometry-based methods can offer separation assurances but face limitations in dense encounters. Recurrent reinforcement learning has therefore been proposed to learn navigation policies that integrate observation-action histories to optimise long-term safety-efficiency trade-offs. However, operational safeguards may require a remote pilot, onboard detect-and-avoid system, or runtime assurance mechanism to temporarily override the learned policy. A research gap remains in understanding how these interruptions affect recurrent policies when the agent regains authority. Benchmark evidence shows that interruptions make task variables less decodable from the recurrent hidden state and increase post-interruption collision risk. The findings suggest that altered recurrent context prolongs post-handback degradation through an action-observation feedback loop, in which changed actions alter the flight path and subsequent recurrent updates. To address this vulnerability, we propose three interruption-aware architectures that condition recurrent processing on interruption context through separate feature streams, dedicated interruption memory, or gated environmental features. Across simulations, all three architectures preserve more decodable task information than a standard recurrent policy. Moreover, the strongest architecture reduces post-interruption collisions by 74% and interruption-induced mission failures by 69% relative to the recurrent baseline, without degrading interruption-free performance. These results show that safe interruption does not guarantee safe handback, and support interruption-aware recurrent design for resilient UAS autonomy.