TranTac: Leveraging Transient Tactile Signals for Contact-Rich Robotic Manipulation

Abstract

Robotic manipulation tasks such as inserting a key into a lock or plugging a USB device into a port can fail when visual perception is insufficient to detect misalignment. In these situations, touch sensing is crucial for the robot to monitor the task's states and make precise, timely adjustments. Current touch sensing solutions are either insensitive to detect subtle changes or demand excessive sensor data.

Here, we introduce TranTac, a data-efficient and low-cost tactile sensing and control framework that integrates a single contact-sensitive 6-axis inertial measurement unit within the elastomeric tips of a robotic gripper for completing fine insertion tasks. Our customized sensing system can detect dynamic translational and torsional deformations at the micrometer scale, enabling the tracking of visually imperceptible pose changes of the grasped object. By leveraging transformer-based encoders and diffusion policy, TranTac can imitate human insertion behaviors using transient tactile cues detected at the gripper's tip during insertion processes.

Demo Video

TranTac performing tactile-guided insertion of a USB plug and a metal key under initial misalignment (1–3 mm). The robot uses only transient tactile signals from fingertip IMUs for real-time 6-DoF pose correction.

Key Results

• ✅ 79% average success rate on visuotactile insertion tasks (outperforms vision-only and force-augmented baselines)
• ✅ 88% success rate in tactile-only insertion with 1–3 mm initial misalignment
• ✅ Generalizes to unseen objects (USB plug, metal key) with ~70% success rate
• ✅ Ultra-low cost ($5 per sensor), compact size (11×11×8 mm), and high bandwidth (3500 Hz)

BibTeX

@article{wu2025trantac,
  title={TranTac: Leveraging Transient Tactile Signals for Contact-Rich Robotic Manipulation},
  author={Wu, Yinghao and Hou, Shuhong and Zheng, Haowen and Li, Yichen and Lu, Weiyi and Zhou, Xun and Shao, Yitian},
  journal={arXiv preprint arXiv:2509.16550},
  year={2025}
}