SenseGlove R1

R1 is a high-precision haptic glove designed specifically for teleoperating dexterous humanoid robot hands. With a total of 40 tracked finger degrees of freedom, integrated grasp force sensing, active force feedback on the fingers, and vibrotactile feedback in the fingertips and palm, the glove enables intuitive and reliable control. It supports both direct teleoperation and imitation learning, making it a powerful interface for advanced teleoperation and robotics systems.

Applications

• Teleoperation of robotic hands
• Imitation learning including force data
• Research in dexterous manipulation

Key Features

• Active Force Feedback (AFFB)
• Thumb, index, middle, and ring fingers are equipped with AFFB modules that provide targeted resistance on the operator’s fingers, simulating rigid objects in the robot’s grasp or the restorative pull of elastic materials. The feedback is strong enough to simulate firm grip resistance.

Vibration Feedback

Vibrotactile actuators are integrated in all fingertips (5) and the palm (3) to allow the operator to feel surface textures such as rough or smooth. They can also convey tactile cues such as button clicks, tool vibrations, light touches, or UI signals.

Precision Finger Tracking

40 rotary Hall-effect sensors (5 fingers × 8 joints) provide robust, millimeter-precision tracking of all fingers at the fingertips and along each phalanx.

Force Sensing

Each AFFB module includes a force sensor, enabling operation and imitation learning with both precise finger motion and grasp force data. Combined with robot hands featuring force sensing or current measurement, it allows highly realistic and stable grasp simulations.

Specifications

• Sensors: 40 × ASM AS5048A 14-bit rotary position sensors; 4 × force sensors integrated in the AFFB module (range: 10g–2500g)
• Actuators: 4 × Active Force Feedback Modules (motor, gearhead, custom pulley system); Motor: Faulhaber 2619S012SR brushed DC; Gearhead: 8:1 spur gear; Mechanical time constant: 6.8 ms; No-load speed: 5900 min⁻¹; Max torque: 9 mNm (continuous), 30 mNm (intermittent)
• 8 × Linear Resonant Actuators; Rise time: 12 ms; Fall time: 55 ms; Resonant frequency: 170 Hz; Vibration force: 1.8 G
• Force Transmission: Tendon-based system routed through articulated finger exoskeleton
• Finger Tracking: 8 DOF per finger (5 fingers), total 40 DOF
• Feedback Types: Force: 1 DOF per finger active pulling force; Vibration: 8 locations (5 fingertips + 3 palm)
• Microprocessor: Raspberry Pi RP2350B
• Connection & Power: USB-C via wired linkbox to PC and wall socket; USB 2.0 with USB PD; Max power draw: 45 W; Wireless: None
• Software Support: Python API, ROS 2
• Glove Sizes: Swappable textile base glove: Small / Medium / Large
• Weight: TBD
• Latency & Performance: Rembrandt → PC latency: ~10 ms; FPS: ~1 kHz
• Supported OS: Windows, Linux

Scope of delivery

• 1 × Left R1 glove with Softglove (size selected)
• 1 × Right R1 glove with Softglove (size selected)
• 1 × Linkbox
• 1 × USB power adapter
• 4 × USB-C cables
• 1 × Left mechanical interface to tracking solution (option selected)
• 1 × Right mechanical interface to tracking solution (option selected)