Look as You Leap: Planning Simultaneous Motion and Perception for High-DOF Robots
Look as You Leap: Planning Simultaneous Motion and Perception for High-DOF Robots
This paper introduces PS-PRM, a GPU-parallelized perception-score-guided probabilistic roadmap planner that jointly handles motion planning and continuous perception constraints (e.g., object detection, human activity recognition) for high-degree-of-freedom robots in static and dynamic environments. The method uses a neural surrogate model to approximate perception scores, integrates them into roadmap search, and enables efficient online replanning. Experiments in simulation and on real robots show that PS-PRM outperforms reinforcement learning and trajectory-optimization baselines.
The authors present a GPU-parallelized probabilistic roadmap planner (PS-PRM) that jointly optimizes motion and perception for high-DOF robots, using a neural surrogate model to approximate perception scores; results are reported in simulation and real-robot experiments but do not claim deployment readiness or safety certification.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
- Environment
- SimulationOpen source S6
- paper: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
nto roadmap planning, and leverages GPU parallelism for efficient online replanning. We demonstrate that our planner outperforms RL- and trajectory-optimization-based baselines in static and dynamic environments in simulation and real-robot experiments.
Open source S6
Source：arXiv Robotics — research abstracts · arxiv.org