An AI‐Powered, All‐Printed, Scalable, Stretchable Triboelectric E‐Skin for Multifunctional Perception in Dexterous Hand

Abstract

ABSTRACT Achieving human‐like dexterity in robotic hands requires electronic skins (E‐skins) that seamlessly integrate multifunctional sensing, decision‐making, and interactive control. However, existing E‐skins for dexterous hands remain limited to single sensing modalities and face scalability challenges due to complex manufacturing processes. Here, we present a triboelectric E‐skin (TE‐Skin) that overcomes the above limitations via an interfacial compression‐assisted coaxial printing technique. This approach enables the scalable fabrication of ultra‐thin sensory arrays that conformably integrate with the entire robotic hand—fingertips, palm, and dorsum. The TE‐Skin simultaneously enables tactile pressure mapping, dynamic trajectory recognition, material discrimination, secure user authentication, and gesture‐based control. Crucially, deep learning algorithms decode complex triboelectric signals, allowing the system to achieve over 95% accuracy in material recognition and user identification. By merging scalable manufacturing and multifunctional sensing, this work provides a versatile platform for next‐generation robotic manipulation and natural human–robot interaction.