analiza aiqbench
This research introduces a hybrid planning-learning architecture that enables Unmanned Underwater Vehicles to navigate dynamically using only onboard sonar and depth data without external maps. By integrating reinforcement learning with global clearance-aware planners and uncertainty-calibrated distillation, the framework significantly enhances safety and robustness in partially observable environments. The study provides a reproducible methodology for building autonomous systems capable of handling complex underwater dynamics through advanced machine learning techniques.
Unified Planning-Learning Framework for Robust UUV Navigation Under Partial Observability
arXiv:2608.05365v1 Announce Type: new Abstract: This paper presents an observation-only autonomy framework for Unmanned Underwater Vehicles (UUVs) navigation in dynamic underwater environments that integrates persistent occupancy mapping, global clearance-aware planning, and risk-aware local control. The proposed pipeline constructs occupancy maps solely from onboard sonar and depth image observations, adapts a clearance-constrained global planner (GP) to provide long-horizon structure, and integrates a reinforcement learning (RL) policy to handle short-range tracking and reactive avoidance. To further support decision-making under partial observability, the system learns a compact latent state representation from onboard sensor data, encoding environmental structure, obstacle dynamics, and uncertainty. Behavior tree (BT) distillation with staged supervision is introduced to improve safety and training stability, while an uncertainty-calibrated distillation mechanism reweights teacher guidance using online latent-model uncertainty, emphasizing uncertain regimes during learning, with time-to-collision (TTC) and clearance cues remaining explicit in planning and local policy features. To demonstrate the efficacy of the framework, a reproducible multi-seed evaluation protocol is established in high-fidelity GPU-accelerated simulation using NVIDIA Isaac Sim, and performance is benchmarked against BT-only and standard RL baselines. The results obtained demonstrate improved robustness and safety under dynamic conditions, thus providing a general pipeline with a unified hybrid planning learning architecture and a reproducible methodology for robust UUV autonomy under partial observability.
Abstract:This paper presents an observation-only autonomy framework for Unmanned Underwater Vehicles (UUVs) navigation in dynamic underwater environments that integrates persistent occupancy mapping, global clearance-aware planning, and risk-aware local control. The proposed pipeline constructs occupancy maps solely from onboard sonar and depth image observations, adapts a clearance-constrained global planner (GP) to provide long-horizon structure, and integrates a reinforcement learning (RL) policy to handle short-range tracking and reactive avoidance. To further support decision-making under partial observability, the system learns a compact latent state representation from onboard sensor data, encoding environmental structure, obstacle dynamics, and uncertainty. Behavior tree (BT) distillation with staged supervision is introduced to improve safety and training stability, while an uncertainty-calibrated distillation mechanism reweights teacher guidance using online latent-model uncertainty, emphasizing uncertain regimes during learning, with time-to-collision (TTC) and clearance cues remaining explicit in planning and local policy features. To demonstrate the efficacy of the framework, a reproducible multi-seed evaluation protocol is established in high-fidelity GPU-accelerated simulation using NVIDIA Isaac Sim, and performance is benchmarked against BT-only and standard RL baselines. The results obtained demonstrate improved robustness and safety under dynamic conditions, thus providing a general pipeline with a unified hybrid planning learning architecture and a reproducible methodology for robust UUV autonomy under partial observability.
8 pages, 6 figures. Accepted for presentation at the 2026 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2026), Philadelphia, PA, USA
Robotics (cs.RO)
arXiv:2608.05365 [cs.RO]
(or arXiv:2608.05365v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2608.05365 arXiv-issued DOI via DataCite (pending registration)
Text integral sub licența deschisă a publicației · 245 cuvinte
Subiecte conexe
Digital States 2026: AI Moves From Experiment to Infrastructure
govtech.com · 24 sept. 2026
News
Quantum Computing Report · 25 iul. 2026
News
Quantum Computing Report · 7 iun. 2026
GuideWalk: Learning Unified Autonomous Navigation and Locomotion for Humanoid Robots across Versatile Terrains
arXiv · 16 iun. 2026