Closing the Reality Gap: Zero-Shot Sim-to-Real Deployment for Dexterous Force-Based Grasping and Manipulation
2026-07-06 • Robotics
Robotics
AI summaryⓘ
The authors developed a way to train robot hands with five fingers to handle objects like humans do, using only computer simulations before trying on real robots. They use detailed touch sensing and estimate joint forces without extra sensors to help the robot understand how it’s touching and moving objects. Their method includes simulating touch quickly, calibrating torque from motor currents, and modeling motor behavior with randomness to make training realistic. They trained the robot to control grip strength and turn objects reliably without extra adjustments after moving from simulation to the real hand.
dexterous robotic handsim-to-real transferreinforcement learningtactile feedbackjoint torque sensingactor-critic PPOmotor current calibrationactuator dynamicsgrasp force controlrobotic manipulation
Authors
Zhe Zhao, Zhibin Li, Yilin Ou, Mengshi Qi
Abstract
Human-like dexterous hands with multiple fingers offer human-level manipulation capabilities but remain difficult to train the control policies that can deploy on real hardware due to contact-rich physics and imperfect actuation. We present a sim-to-real reinforcement learning method that leverages dense tactile feedback combined with joint torque sensing to explicitly regulate physical interactions. To enable effective sim-to-real transfer, we introduce (i) a computationally fast tactile simulation that computes distances between dense virtual tactile units and the object via parallel forward kinematics, providing high-rate, high-resolution touch signals needed by RL; (ii) a current-to-torque calibration that eliminates the need for torque sensors on dexterous hands by mapping motor current to joint torque; and (iii) actuator dynamics modeling with randomization to account for non-ideal torque-speed effects and bridge the actuation gaps. Using an asymmetric actor-critic PPO pipeline, we train policies entirely in simulation and deploy them directly to a five-finger hand. The resulting policies demonstrate two essential human-hand skills: (1) command-based controllable grasp force tracking and (2) reorientation of objects in the hand, both of which are robustly executed without fine-tuning on the robot. By combining tactile and torque in the observation space with scalable sensing and actuation modeling, our system provides a practical solution to achieve reliable dexterous manipulation. To our knowledge, this is the first demonstration of controllable grasping on a multi-finger dexterous hand trained entirely in simulation and transferred zero-shot on real hardware.