A robot with artificial muscles learned to walk in a video game and then instantly did it in real life without needing any practice.
April 13, 2026
Original Paper
Sim-to-Real Transfer for Muscle-Actuated Robots via Generalized Actuator Networks
arXiv · 2604.09487
The Takeaway
Artificial muscles are notoriously difficult to control because they are stretchy and unpredictable, unlike the rigid motors used in most robots. By using a neural network to perfectly mimic these messy physics in a computer, researchers can now train robots safely in virtual reality before letting them loose in the real world.
From the abstract
Tendon drives paired with soft muscle actuation enable faster and safer robots while potentially accelerating skill acquisition. Still, these systems are rarely used in practice due to inherent nonlinearities, friction, and hysteresis, which complicate modeling and control. So far, these challenges have hindered policy transfer from simulation to real systems. To bridge this gap, we propose a sim-to-real pipeline that learns a neural network model of this complex actuation and leverages establis