KORTHOS / ECOSYSTEM / PRODUCT

Uhand

About

About Uhand

Uhand captures hand and finger movement together with high-precision tactile measurements during a manipulation demonstration. The system records contact location and force alongside the visual and motion streams, preserving the physical information that is lost when training data consists only of video.

ChangingTek designed the hand as part of a full multimodal data-collection workflow for humanoid and dexterous-manipulation models. The captured data can be used to train, evaluate and refine robot-hand policies before transferring them into deployed end effectors.

Capabilities

Source-backed capabilities

  • Capability areas: Tactile feedback
  • Workflows: Robot learning data
  • Operating environments: Robotics labs, manipulation research platforms, dexterous-hand testbeds, AI model-training setups, and data-collection rigs.
  • Target markets: Robotics researchers, embodied-AI teams, manipulation labs, and developers building tactile or dexterous-control datasets.
  • Deployment model: End-effector component integrated with robot arms, data pipelines, perception systems, and learning workflows.
  • Multimodal data collection

    Claimed - 2026-06-01 - ChangingTek's product release introduces UHand as a data-acquisition gripper providing tactile sensing for embodied-intelligence data-loop workflows.

    Source
  • Tactile interaction sensing

    Claimed - 2026-06-01 - ChangingTek's product release introduces UHand as a data-acquisition gripper providing tactile sensing for embodied-intelligence data-loop workflows.

    Source

Timeline

  • Changingtek Robotics launches Uhand tactile sensing data-collection hand

    2026-06-01 - Product Launch - Changingtek Robotics globally launched Uhand, an in-house developed tactile data-collection hand for embodied-intelligence research, precision manipulation and algorithm validation. Uhand captures synchronized visual, pose, force and tactile datasets at 30 Hz, with a tactile-array resolution of 2.34 taxels per square centimetre. The product supports 0-160 N force detection, 0.1 N force precision, 0.7 mm XYZ positioning accuracy and up to four hours of battery runtime in a 600 g device.

    Source

Context

Market context

Uhand is Changingtek Robotics' tactile data-collection hand for embodied-intelligence research, precision manipulation, and robot-learning data capture.

  • Target environment: Robotics labs, manipulation research platforms, dexterous-hand testbeds, AI model-training setups, and data-collection rigs.
  • Workflow context: Tactile data capture, force and pose sensing, manipulation dataset collection, algorithm validation, and contact-rich robot-learning workflows.
  • Customer context: Robotics researchers, embodied-AI teams, manipulation labs, and developers building tactile or dexterous-control datasets.
  • Deployment model: End-effector component integrated with robot arms, data pipelines, perception systems, and learning workflows.
  • Commercial maturity: Newly launched product positioned around tactile sensing and synchronized multimodal data collection.
  • Adoption constraints: Adoption depends on sensor fidelity, hand durability, robot-arm compatibility, calibration workflow, dataset tooling, and model-training integration.
  • Market position: Tactile end-effector component for robot-learning and manipulation research.
  • Adjacent products: Mini45 Force/Torque Sensor, OAK-D Pro AI Stereo Camera, AK80-9 V3.0 Robotic Actuator

Related Systems