01
Stereo vision
Two synchronized cameras track visual features and provide motion and scale constraints.
Infrastructure-independent spatial tracking
Track people and objects without external beacons
The M-One 6-DoF Tracker estimates position and orientation directly on the device by combining stereo vision with high-rate inertial measurements.
Calculated on-device
Six degrees of freedom from two complementary signals
A live visual-inertial odometry pipeline fuses synchronized camera and IMU measurements into a continuous pose stream.
02
Inertial sensing
An onboard ICM45686 IMU supplies gyroscope and acceleration data at approximately 200 Hz in the tested prototype.
03
Live 6-DoF pose
The estimator outputs timestamped 3D position and quaternion orientation at approximately 30 Hz in the tested configuration.
Adaptable mounting
Put spatial context where the action happens
Wear the tracker on the chest, hand, or leg, or attach it to tools, boxes, and other equipment. Strap-based adapters and mechanical mounting points support fast setup across different forms and materials.
- Chest-mounted person tracking
- Object and tool tracking
- Industrial XR training
- Contextual object motion for humanoid data
XR interoperability
One OpenXR-based integration path
The reference Unity integration uses OpenXR for session lifecycle, spaces, poses, and input. The prototype was demonstrated with HTC Vive Focus Vision and Meta Quest 3.
Portable coreTracker pose, calibration logic, and XR interaction retain one overall architecture.
Device layerPassthrough and hand tracking depend on the capabilities of the selected headset.
Prototype evidence
Validated across realistic integration scenarios
The documented M4 evaluation covered robot-mounted, human-held, human-worn, multi-headset, lighting, occlusion, and feature-density scenarios.
0.95°Measured rotational accuracy
1.54 cmMeasured translational accuracy under realistic, dynamic conditions
2 headsetsHTC Vive Focus Vision and Meta Quest 3 demonstrations
No delayObserved pose-stream delay in the simulated occlusion test
These are prototype evaluation results, not universal product specifications. The report also identifies limits during fast motion and in visually feature-poor environments.
Rotational accuracy
Robot-mounted tracking
See the recorded comparison of tracker and robot rotation values for roll, pitch, and yaw.
Watch demonstration →Object tracking
Three loops, one return point
See a human-held tracker use loop closure while moving around an office environment.
Watch demonstration →Add spatial context
Track what moves without rebuilding the room
Tell us what must be tracked, where it moves, and which XR or data workflow will consume the pose stream.