DexSkin: High-Coverage Conformable Robotic Skin for Learning Contact-Rich Manipulation

2025-09-23

AI Summary

To address the challenge of insufficient high-coverage, adaptive tactile sensing for dexterous robotic manipulation in contact-rich tasks, this paper introduces DexSkin—a stretchable capacitive electronic skin. DexSkin employs a multi-layer, elastomeric architecture to achieve full-fingertip tactile coverage with high sensitivity (<0.1 nF/mm²) and enables calibration-free transfer across geometric deformations and sensor instances. Integrated with a learning-from-demonstration framework and online calibration strategy, it is deployed on a parallel-jaw gripper to deliver real-time tactile feedback. On a physical robot platform, DexSkin successfully enables online reinforcement learning to perform challenging tasks—including object flipping and rubber-band winding—demonstrating strong efficacy and generalization in data-driven tactile manipulation. The core contribution is the first realization of a closed-loop system combining full-fingertip adaptive capacitive sensing with model-transferable tactile learning for dexterous manipulation.

Technology Category

• Intelligent Robots: Manipulation
• Machine Learning: Learning with Manifolds
• Humans and AI: Interaction Techniques and Devices

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Human skin provides a rich tactile sensing stream, localizing intentional and unintentional contact events over a large and contoured region. Replicating these tactile sensing capabilities for dexterous robotic manipulation systems remains a longstanding challenge. In this work, we take a step towards this goal by introducing DexSkin. DexSkin is a soft, conformable capacitive electronic skin that enables sensitive, localized, and calibratable tactile sensing, and can be tailored to varying geometries. We demonstrate its efficacy for learning downstream robotic manipulation by sensorizing a pair of parallel jaw gripper fingers, providing tactile coverage across almost the entire finger surfaces. We empirically evaluate DexSkin's capabilities in learning challenging manipulation tasks that require sensing coverage across the entire surface of the fingers, such as reorienting objects in hand and wrapping elastic bands around boxes, in a learning-from-demonstration framework. We then show that, critically for data-driven approaches, DexSkin can be calibrated to enable model transfer across sensor instances, and demonstrate its applicability to online reinforcement learning on real robots. Our results highlight DexSkin's suitability and practicality for learning real-world, contact-rich manipulation. Please see our project webpage for videos and visualizations: https://dex-skin.github.io/.

Problem

• Replicating human skin's tactile sensing capabilities for robotic manipulation systems
• Providing sensitive localized tactile sensing coverage across entire robotic finger surfaces
• Enabling learning of contact-rich manipulation tasks requiring full finger sensing coverage

Innovation

• Soft conformable capacitive electronic skin
• Provides full-finger tactile coverage for manipulation
• Enables calibration for model transfer across sensors