Autonomous Particle Image Velocimetry Using Zero-Shot Sim-to-Real Reinforcement Learning
Autonomous Particle Image Velocimetry Using Zero-Shot Sim-to-Real Reinforcement Learning
This research presents a method where a reinforcement learning agent controls a six-degree-of-freedom manipulator equipped with a camera to autonomously position itself for particle image velocimetry (PIV) measurements. Trained entirely in a photorealistic simulation using computational fluid dynamics data, the agent achieves zero-shot transfer to a real water tunnel. It successfully locates and frames hydrofoil wakes generated by varying profiles and angles of attack, resolving flow fields at higher resolution than fixed camera views while reducing measurement time and image pair requirements.
The study demonstrates a reinforcement learning agent trained in simulation that successfully transfers to a real water tunnel to optimize camera placement for particle image velocimetry, achieving higher resolution flow measurements with fewer image pairs than fixed-camera baselines.
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 S3
- paper: not_reported
- code: not_reported
- weights: not_reported
- dataset: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
forcement learning agent to control a 6-DoF manipulator equipped with a camera, enabling it to locate, frame and resolve a hydrofoil wake in a water tunnel. The agent commands the camera pose and observes the scene only through the velocity field and uncertainty estimated from time-separated image pairs. It is trained entirely in a photorealistic simulation of our water tunnel, with wakes generated by Computational F
Open source S3
luid Dynamics (CFD) simulations of hydrofoils with varying profiles, positions and angles of attack. After zero-shot transfer to the real water tunnel, the policy reliably locates and frames the wakes of randomly placed hydrofoils. It resolves the wake at a higher resolution than a fixed camera view and with lower error in the measured flow statistics, in roughly half the time of a full scan and with far fewer image
Open source S4
Source：arXiv Robotics — research abstracts · arxiv.org