Collect, train and deploy
Choose your solution. Capture anywhere, across every kind of movement: from walking and running to jumping, crawling, and climbing stairs. Start with the core, then expand with the add-ons your application needs.
Why it works for humanoid training
Xsens’ patented sensor-fusion IP closes the data gaps other capture tools leave open: full magnetic immunity, ground-truth foot contact and a validated biomechanical model no one else can match. The result is deployment-grade data you can capture anywhere your humanoid will operate.
Bridge the gap between training and deployment environments
Use it anywhere. Captures research-grade fidelity wherever the work happens: indoors, outdoors, on factory floors, warehouses, fields and construction sites. No cameras, no markers, no controlled lighting, so you train the policy where the humanoid will actually deploy.
Integrate in your pipeline
Standard exports and native integrations with the simulators, middleware and animation tools humanoid teams already use. Data lands in your training stack ready to drive policy training, not waiting for conversion.
Less time in simulation
Patented sensor-fusion and a full kinematic chain mean policies learn from physically correct ground truth, not noisy proxies. High-end data needs fewer synthetic hours to generalise, compressing the sim-to-real gap and the synthetic-augmentation budget.
Tailored for any body
First inertial system with native male and female anatomical models (Biomech 2.0). Segment lengths, hip dimensions and centers of mass scale to each operator. Datasets reflect the diversity of real-world humans.
Consistent data stream, no interference
The only inertial mocap with full magnetic immunity. Steel rebar, electric motors and automotive lines do not corrupt the signal. No constant recalibration. Capture clean data inside the factories your humanoid will actually deploy in.
Humanoid balance
More foot contact points than any other inertial system, plus full Center of Pressure and Center of Mass estimation. The robot inherits ground-truth contact phase, not heel-toe approximation, so balance and gait policies stand up to real terrain, including slopes, inclines and uneven ground.
Two core solutions for robot training
Teleoperation and Live Control
Stream human motion to a robot in real time, and record every demonstration as training data in the same session.
Human Data Collection
Capture human tasks anywhere and process them into training-ready datasets for AI/ML, simulation and robot control pipelines.
Control and train your humanoid in real time
Wear the suit, stream motion straight to the robot, and record every demonstration at the same time. Works over wireless WiFi and over wired tethered connections, so the setup fits any environment from a research lab to an industrial teleop cell.
Collect Human Movement Data in the Wild
Capture humans doing real work wherever the work happens, indoor or outdoor, on factory floors, in warehouses, outside in the field. Record the full range of human motion, from walking and running to jumping, crawling and staircase climbing, and turn those sessions into training-ready datasets.
Expand your setup with 3rd-patry tools
Both Live Teleoperation and Human Movement Data Collection share the same optional modules. Layer the ones that match the task you actually want the humanoid to perform.
Integrated finger tracking
Manus or StretchSense gloves layer in full per-finger DOF inside the same capture loop. Powers dexterous manipulation, grasping and contact-rich tasks in live teleop or offline datasets. Learn more
Object tracking
Track tools, payloads and workspace objects alongside the human with HTC Ultimate Tracker or Vive. The dataset captures the interaction, not just the body. Learn more
Video sync
Time-locked multi-camera video streamed alongside motion. Vision and biomechanics on the same clock, ready for VLA training and annotation.
HMD integration
Pair captures with VR headset data for first-person datasets, immersive review and full-body teleoperation. Meta Quest and Pico VR support coming soon.
Absolute positioning
Lock the suit to a global reference frame with HTC Ultimate Tracker or Vive. Datasets become spatially grounded for interaction and scene-aware tasks. Learn more
Don't want to build it yourself? Discover data Collection as a Service
For teams that want the dataset without building the data-collection operation. Temporarily install the Xsens-Robotics infrastructure on-site, run the sessions to your spec, hand over the dataset, and make sure you can use it in your training pipeline. Scoped, time-boxed, high-end data. We can connect you to partners who run data collection at scale.
Ready to pick your solution?
Tell us what you want to build. We will configure the right package and add-ons for your humanoid program.