SenseGlove R1
An active force-feedback system with a separate Python API and ROS wrapper. Check prototype conditions, message units and external wrist tracking.
DROID RESEARCH · SOURCES CHECKED
Can Nova 2 software and tracking assumptions be carried over to R1?
R1 has a separate Python API and ROS wrapper. Its developer reference describes active finger force and external wrist tracking. Use the R1 interfaces and record the prototype revision; Nova 2’s Jazzy instructions are not proof of R1 distribution support.
What the sources establish
Manufacturer documentation is attributed below. These values are not independent measurements by Droid.
| Field | Reported value | Conditions and evidence |
|---|---|---|
| Finger force feedback | Active force on thumb through ring finger | No pinky force feedback; active force is distinct from Nova 2’s passive brakes. SenseGlove ↗ |
| Vibration feedback | Fingertip and palm actuators | Separate from the finger force mechanism. SenseGlove ↗ |
| Raw exoskeleton update rate | 1000 Hz without filtering | Raw linkage data; not a guaranteed application rate or human-fingertip accuracy. SenseGlove ↗ |
| Global wrist position | External tracking required | The glove cannot determine its own world position. SenseGlove ↗ |
| ROS 2 interface | Dedicated R1 Python API wrapper | Pinned commit 5005aff. Specific ROS distribution support is not established by the reviewed setup guide. SenseGlove ↗ |
| ROS message units | Angles: rad; fingertip positions: mm; forces: mN | Fingertip positions are relative to each finger-base origin; preserve both frame and units. SenseGlove ↗ |
| Python API host platforms | Windows and Linux listed | USB-C required. Exact processor, OS and API release still need validation. SenseGlove ↗ |
| Tracking test conditions | API page identifies prototype 0.4 jig tests | Aligned exoskeleton linkages; the page separately discusses prototype 0.5. Do not apply the old measurements to every delivered revision. SenseGlove ↗ |
What this means for an integration
Droid’s analysis of the cited documentation:
- Treat the force-command boundary as an explicit unit conversion. If your robot reports newtons, the R1 message expects milliNewtons: 1 N equals 1000 mN. Check sign, finger order and device limits before enabling output.
- A ROS Point field does not make the documented coordinates meters or world-relative. Convert millimeters only at a defined boundary, and transform from the correct finger-base frame.
- Keep raw, filtered and application rates separate. The API documentation ties throughput to CPU, glove count and GUI state, and describes prototype-specific jitter filters. Measure the complete setup you intend to deploy.
Still to establish
- Delivered prototype or production revision and current availability
- Validated ROS distribution / Python API release pair
- Independent tracking and closed-loop latency measurements
- Configured price, Link Box, tracker and support scope
Before a larger capture run
- Record the glove revision and API / ROS commit before testing.
- Verify units, finger order and coordinate frames with a static fixture.
- Check tracking loss and filter settings with both gloves connected.
- Validate bounded force commands and an operator stop mechanism.
Source record
- R1 developer reference ↗SenseGlove · R1 / formerly Rembrandt; hardware revision unspecified · Checked 2026-11-10
- R1 ROS messages and force commands ↗SenseGlove · Commit 5005aff, October 27, 2026 · Checked 2026-11-10
- R1 ROS initial setup ↗SenseGlove · Commit 5005aff; Windows instructions point to WSL2 · Checked 2026-11-10
- R1 API: tracking conditions and prototype notes ↗SenseGlove · Prototype 0.4 test conditions; 0.5 and API filter revision notes · Checked 2026-11-10
Make the next decision.
START WITH THE WORKFLOWBring us the task.
Request a project review Bring us the task.
Let’s find the fit.
Share your environment, the job to be done, and the outcome you need to measure.