ReTouch: Empowering Contact-Rich Dexterous Manipulation with Online-Refined Tactile Prediction
2026-08-03 • Robotics
Robotics
AI summaryⓘ
The authors developed ReTouch, a system that helps robots use touch signals better to handle objects with careful finger movements. It uses a special way to understand touch called the Tactile-Patch Encoder, which keeps track of which finger feels what. ReTouch also predicts touch and movements ahead of time and updates these predictions as it actually touches objects, helping the robot adjust quickly. They tested ReTouch on seven tricky tasks with a real robot hand and found it worked much better than previous methods. This shows their approach improves how robots can manage delicate, touch-based tasks.
tactile signalsdexterous manipulationtactile perceptionvision-language-action modelTactile-Patch Encoderclosed-loop controlaction predictionrobotic handcontact-rich tasksreal-robot experiments
Authors
Shiqi Zhang, Xin Zhang, Yedong Shen, Jiajun Deng, Yuxuan Gao, Sha Zhang, Yuan Zhang, Kaixue Long, Jiajia Wu, Jia Pan, Yao Li, Yanyong Zhang
Abstract
Fusing tactile signals has proven effective for contact-rich manipulation, enabling robots to perceive contact states and adapt to rapidly changing physical interactions. Yet effectively integrating tactile feedback into dexterous manipulation remains underexplored. In this work, we introduce ReTouch, a vision-language-action model (VLA) that supports contact-rich dexterous manipulation through tactile predictions continually refined online using execution-time feedback. ReTouch builds on two main innovations for tactile representation and closed-loop action generation. First, its Tactile-Patch Encoder represents tactile observations as structured tactile patch features that preserve finger identity and local contact structure, providing contact cues for fine-grained dexterous control. Second, its high-frequency action module jointly predicts future tactile states and action chunks and refines both using incoming tactile feedback during execution. This closed-loop refinement keeps tactile predictions aligned with evolving physical interactions, enabling responsive action correction and improving robustness to contact changes and execution errors. We further introduce XHT-Dataset, comprising 900 real-world demonstrations across seven contact-rich tasks collected on an XHand--UR7e platform, and evaluate ReTouch through closed-loop real-robot experiments. ReTouch surpasses the strongest baseline by 18.4 and 23.8 percentage points in average success rate under standard and challenging conditions, respectively, demonstrating its effectiveness and robustness.