OPERATOR PROFILE // PARTNER REFERENCE DOCUMENT

RT-FUSION
PARTNER INFORMATION

Technical reference for data platform partners, research collaborators, and enterprise buyers evaluating RT-Fusion as a field acquisition operator. Covers rig specifications, European operational coverage, output formats, and onboarding.

01 // WHAT THE RIG CAPTURES

RT-Fusion deploys a dual-sensor rig worn by a human operator in real-world environments. The GoPro Hero 13 — chest-mounted on a custom vibration-isolated rig with continuous weatherproof power — captures wide-angle world-view video at up to 5.3K with 200Hz IMU and GPS telemetry embedded in the GPMF stream. Simultaneously, the Ray-Ban Meta Gen 2 — worn at eye level — captures first-person POV clips (up to 3 minutes each) reflecting true human head orientation and social field of view. Both sensors run in parallel with audio-synced timestamps, producing paired data that no vehicle-mounted or robot-mounted sensor can replicate.

[ WORLD SENSOR ]

KINEMATICS & CONTEXT

GOPRO HERO 13 — CUSTOM ACQUISITION RIG

  • Resolution4K / 5.3K (config dependent)
  • Color10-bit GP-LOG
  • IMU200Hz accel + gyro
  • GPSEmbedded GPMF
  • PowerContinuous weatherproof mag-link
  • MountChest, vibration-isolated
Serves: All use cases (AD, Robotics, DMS, Wearable AI)
[ POV SENSOR ]

INTENT, ATTENTION & POV

RAY-BAN META GEN 2

  • Resolution3K (C1) / 1080p (C2), clips ≤3 min
  • AudioBinaural spatial array
  • POVTrue human eye-level
  • Social cuesMicro-gesture capture
  • Gaze proxyHead-locked center frame — no eye tracking
  • Form factorConsumer smart glasses
Serves: AD (VRU intent), Robotics (demonstrations), DMS (driver POV + head orientation), Wearable AI (form-factor match)

02 // OPERATIONAL COVERAGE

RT-Fusion is based in Venlo, Netherlands, with regular operations across the Netherlands and Belgium. Deployable to Germany, France, Austria and Switzerland within 48 hours' notice. All acquisition is conducted on public infrastructure under applicable data protection law (GDPR in the EU, the FADP in Switzerland). Delivery is raw by default; face and licence-plate anonymisation is available as a paid add-on.

● ACTIVE BASE
Netherlands
Venlo · Amsterdam · Rotterdam · Utrecht
● ACTIVE BASE
Belgium
Brussels · Antwerp · Ghent
○ DEPLOYABLE
Germany
48h notice — Düsseldorf, Cologne, Hamburg, Berlin
○ DEPLOYABLE
France
48h notice — Paris, Lyon, Lille
○ DEPLOYABLE
Austria
48h notice — Vienna, Graz
○ DEPLOYABLE
Switzerland
48h notice — Zurich, Basel, Geneva
// Acquisition capacity: 4h+ continuous GoPro capture per day, plus Ray-Ban POV clips (up to 3 min each). Multi-day campaigns and recurring deployments available. Custom locations outside the above countries considered on request.

03 // OUTPUT & DELIVERY FORMAT

Every acquisition day produces paired, time-synchronized files from both sensors. Assets are structured per scene and indexed by scenario type and sensor configuration. The formats are standard: MP4 loads directly in OpenCV and PyTorch, GPMF telemetry parses with open-source tools (gpmf-parser, gopro2gpx), and conversion to ROS 2 bag takes a short script.

VIDEO
MP4
H.264 / H.265
TELEMETRY
GPMF
GPS + 200Hz IMU
METADATA
JSON
Per scene
DELIVERY
Flexible
Cloud / shared drive / physical

Metadata sample — per-scene JSON structure:

metadata_sample.json
{
  "timestamp_utc": "2026-02-11T09:14:22.045Z",
  "frame_id": 4920,
  "environment": {
      "location": "NL_Amsterdam_Canal_District",
      "weather": "overcast_diffuse",
      "surface": "asphalt_bike_lane"
  },
  "telemetry": {
      "imu_accel_x_y_z": [0.02, -0.81, 0.15],
      "speed_mps": 5.8
  },
  "sensors": {
      "world_cam_file": "GH010492.MP4",
      "attention_cam_file": "RM010492.MP4",
      "head_pose_proxy": true
  }
}

Sample output — inspect before you engage:

[ OPEN FULL SAMPLE PORTFOLIO — 7 FOLDERS, PAIRED SENSORS ]

Compatible with:

PyTorch DataLoader ROS 2 bag (via conversion) OpenCV NVIDIA Isaac (via conversion) gpmf-parser / gopro2gpx (GPMF parse) Custom pipelines

04 // ONBOARDING

The typical engagement starts with a brief scoping call or written brief — target use case, required scenarios, volume, and timeline. RT-Fusion works to a per-acquisition-day model: you specify what you need, RT-Fusion deploys, captures, and delivers structured assets within the agreed window. For platform partners running repeat acquisition programmes, a standing brief and scheduling cadence can be agreed upfront.

What to include in your brief:

  • ›Target use case (AD, Robotics, DMS, Wearable AI, or mixed)
  • ›Scenario requirements — environment type, activities, conditions
  • ›Volume — number of acquisition days or hours of footage
  • ›Preferred locations or geographic constraints
  • ›Delivery timeline and format preferences
  • ›Anonymisation requirements (face / plate blurring)

Pricing reference:

Per acquisition day €1,500 – €2,500
Multi-day campaigns Quoted per project
Standing programmes Retainer available
Delivery MP4 + GPMF + JSON

Send your brief or intake process:

Prefer email? [load email]  ·  PGP key available