Real-world data for physical intelligence
From real-world egocentric capture to model-ready training data.
The data robots learn from.
We capture real work as it happens and deliver robotics-ready training data: enriched, annotated, in your format.
Real work, captured where it happens.
From kitchens and workshops to retail and production floors—real-world egocentric data collected through our global network.
Five layers. One system.
How a worker's shift becomes training data: one pipeline from photons to your schema, with quality measured at every step. Then the loop closes: your eval results become the next capture brief.
See how raw capture becomes model-ready dataCapture
Global shutter rigs worn during real shifts. 20,000 collectors across five continents.
Sync
Timestamps written into the pixels of every frame. Sync survives every copy, transcode, and cut.
Refine
Automated ingest: hardware transcode, hand pose and depth enrichment, session assembly.
Verify
Every batch ships with a QC report: frame timing, drop rate, sync integrity. Under 0.1 percent frames lost, measured.
Deliver
Your schema, not ours: mcap, custom protobuf contracts, per-customer format compilers. Fleet firmware ships over the air.
Every batch graded. Weak batches never ship.
Task claims are checked against pixels, clip by clip. Every delivery carries manifests, checksums, and a batch grade.
Vision, motion, and sound. One clock.
The rig our own collectors wear every day, available to labs and data teams. Field problems become hardware revisions in weeks, not quarters. Ultra-wide field of view: arms overhead, hands still in frame.
- Cameras
- 2× global shutter, color, 1920×1200, up to 60 fps
- Stereo sync
- <1 µs
- IMU to exposure
- 6-axis, 300+ Hz, timestamped in every frame, <10 µs
- Audio input
- 2× onboard microphones, stereo, synchronized with video
- Multi-device sync
- <100 µs across additional USB cameras
- Field of view
- Up to 200° diagonal / 110° vertical fisheye; rectified FOV fixed per delivery profile
- Calibration
- Measured per-device report: stereo extrinsics + camera-IMU
- Rectification
- Per-device calibration applied before writing: fisheye distortion removed, so every mcap on the card is training-ready as recorded
- Output
- mcap for ROS 2, or synchronized raw video + IMU + stereo audio streams
- Runtime
- 12+ hours on a power bank
- Live preview
- Wi-Fi live view on a phone to confirm framing before and during recording, so less footage is wasted
- Storage
- 64 GB internal SSD, so it records even without a card; 128 GB high-speed microSD included
- Capacity
- About 24 hours of stereo H.265 per 128 GB card, roughly a week of field sessions
- Configurations
- Stereo headband, multi-camera headband, wrist camera
- Video encoding
- H.265 / HEVC, GOP 30 = 1 s keyframe interval at 30 fps
- Frame types
- I and P frames only, no B-frames
In production use with US customers. In stock: $1,180 per unit, $850 at 25 or more. Large batches ship in 1 to 2 weeks.
Tell us what you're training.
We scope the dataset, design the pipeline, and deliver training-ready data on your timeline. Start with a sample plan and the smallest useful pilot.