# Embedding high-resolution touch across robotic hands enables adaptive human-like grasping

arXiv: https://arxiv.org/abs/2412.14482 · HTML: https://mlsanity.com/paper/2412.14482/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 19 Dec 2024.

## Built on by (most cited first, 8 of 8)

- [Taccel: Scaling Up Vision-based Tactile Robotics via High-performance GPU Simulation](https://mlsanity.com/paper/2504.12908/) (2025)
- [Simultaneous Tactile-Visual Perception for Learning Multimodal Robot Manipulation](https://mlsanity.com/paper/2512.09851/) (2025)
- [TacCoRL: Integrating Tactile Feedback into VLA via Simulation](https://mlsanity.com/paper/2606.11743/) (2026)
- [Blind Dexterous Grasping via Real2Sim2Real Tactile Policy Learning](https://mlsanity.com/paper/2606.11767/) (2026)
- [Adaptive Vision-Language Grasping via Composable Foundation Priors and Generalizable Grasp Synthesis](https://mlsanity.com/paper/2609.04096/) (2026)
- [Scalable Dexterous Robot Learning with AR-based Remote Human-Robot Interactions](https://mlsanity.com/paper/2602.07341/) (2026)
- [TrackDeform3D: Markerless and Autonomous 3D Keypoint Tracking and Dataset Collection for Deformable Objects](https://mlsanity.com/paper/2603.17068/) (2026)
- [Let Robots Feel Your Touch: Visuo-Tactile Cortical Alignment for Embodied Mirror Resonance](https://mlsanity.com/paper/2605.14571/) (2026)

## Abstract (arXiv)

Developing robotic hands that adapt to real-world dynamics remains a fundamental challenge in robotics and machine intelligence. Despite significant advances in replicating human hand kinematics and control algorithms, robotic systems still struggle to match human capabilities in dynamic environments, primarily due to inadequate tactile feedback. To bridge this gap, we present F-TAC Hand, a biomimetic hand featuring high-resolution tactile sensing (0.1mm spatial resolution) across 70% of its surface area. Through optimized hand design, we overcome traditional challenges in integrating high-resolution tactile sensors while preserving the full range of motion. The hand, powered by our generative algorithm that synthesizes human-like hand configurations, demonstrates robust grasping capabilities in dynamic real-world conditions. Extensive evaluation across 600 real-world trials demonstrates that this tactile-embodied system significantly outperforms non-tactile-informed alternatives in complex manipulation tasks (p<0.0001). These results provide empirical evidence for the critical role of rich tactile embodiment in developing advanced robotic intelligence, offering new perspectives on the relationship between physical sensing capabilities and intelligent behavior.