Event
Proposes three interruption-aware recurrent reinforcement learning architectures for UAS navigation
1 reports · 1 sources
Christopher Conrad proposes three interruption-aware recurrent reinforcement learning architectures for scenarios in which unmanned aerial systems resume autonomous navigation after temporary takeover by teleoperation or safety mechanisms, incorporating interruption context into recurrent processing via independent feature streams, dedicated interruption memory, or gated environmental features. The paper states that in simulation all three architectures retain more decodable task information than a standard recurrent policy; the strongest architecture reduces post-interruption collisions by 74% and interruption-induced task failures by 69% relative to the recurrent baseline, without harming performance in the absence of interruptions.
Report timeline
- Interruption-aware recurrent reinforcement learning improves UAS navigation safety for temporary overrides
To address the increased collision risk when unmanned aircraft systems (UAS) resume autonomous navigation after temporary takeover by teleoperation or safety mechanisms, the study proposes three interruption-aware recurrent reinforcement learning architectures that incorporate interruption context into recurrent processing through independent feature streams, dedicated interruption memory, or gated environmental features.
arXiv cs.RO ·