Asv control policies trained in simulation transfer directly to real boats
Sim-to-Real RL for ASVs using SysID
Robotics
Summary
Controlling autonomous boats in changing water conditions is hard, and teaching them using computer simulations usually takes a lot of detailed water and vehicle data. The authors created a system that only needs a simple 3D model of the boat and some short real-world trial runs to estimate how the boat moves and its engine power. They then train control strategies entirely in simulation, which work on the real boat right away without extra tuning. This makes it easier and faster to develop smart controllers for autonomous surface vehicles.
What this means in practice
- •For marine robotics engineers: Train ASV control policies in simulation without detailed hydrodynamic data and deploy them directly on real boats for navigation and station keeping.
- •For autonomous vehicle software developers: Build and refine vehicle dynamics models from limited field data and CAD models to enable sim-to-real RL for surface vehicle control.
Authors
Cody Sheltraw, Tsimafei Lazouski, Maani Ghaffari, Alan Papalia
Abstract
Autonomous Surface Vehicles (ASVs) operating in dynamic marine environments require robust control policies for tasks such as path following and station keeping, making reinforcement learning (RL) a promising alternative to classical controllers. However, existing ASV simulators rarely support parallel environments for RL training. Such existing simulators require accurate hydrodynamic modeling from computational fluid dynamics solvers or towing tank tests for setting hydrodynamic parameters to address the sim-to-real gap. To address these challenges, we present an ASV simulator and accompanying pipeline that enables training policies starting from unknown vehicle dynamics. Our framework uses only a CAD model and brief set of open-water field trajectories for approximating and refining both hydrodynamic and thruster parameters. Real-world deployments on a BlueBoat ASV demonstrate successful zero-shot sim-to-real transfer in path following and station-keeping tasks without prior hydrodynamic and propeller information.