Low fidelity simulator enables robotic fish control in real water
All You Need Is Low Fidelity: Zero-Shot Sim-to-Real of Learned Robotic Fish Control
Robotics
Summary
Controlling underwater robots like robotic fish is usually very hard because simulating water movement accurately takes a lot of computer power. The authors show that a very simple simulator with rough water physics is enough to teach a robotic fish how to swim and follow targets. Their robot uses only simple sensors and runs the same control program in a real pool without any changes. This approach suggests you don’t need perfect simulation to control underwater robots well.
What this means in practice
- •For robotic control engineers: Use simple fluid models in simulation to develop controllers that work directly on underwater robots without extra tuning.
- •For aquatic robotics developers: Deploy closed-loop fish-like robots outdoors for tasks like target tracking and disturbance handling with controllers learned from minimal simulations.
Authors
Liam Maloney, Simon Ramchandani, Mike Y. Michelis, Ronan Hinchet, Robert K. Katzschmann
Abstract
Complex tasks for underwater robots remain limited by the capabilities of their controllers. Learning a better one for a soft, underactuated robotic fish trades simulator cost against fidelity. We show that an intentionally low-fidelity simulator is enough: a stateless, quasi-steady fluid model with no wake and no added-mass history suffices to learn a \emph{general}, closed-loop controller that transfers to hardware without tuning. Our platform is a soft, single-motor, tendon-driven fish whose policy observes only what the hardware can measure. A staged pipeline grounds the simulator in two independent identifications, fixing the tail dynamics and a stateless fluid model; the policy then acts through a band-limited rhythmic trajectory generator rather than commanding the tail directly. Deployed unchanged in an outdoor pool, a single policy performs closed-loop target reaching, disturbance rejection, and out-of-distribution target acquisition and tracking. The transfer rests on the constraint rather than the fidelity: the generator cannot leave the band over which the fluid was identified. This raises the question of how much of the physics can reside in the controller rather than in the simulator.