First ever portable motion capture system.
Deploy anywhere at anytime
For athletes, coaches, creators, artists, researchers. More affordable and easy to use.
Robotics & AI Training Data
Ground-truth dataHigh-fidelity multi-camera pose datasets for imitation learning, sim-to-real validation, and perception model training — captured, not simulated.
Clinical & Research
Motion dataLab-grade accuracy with on-device processing — no patient data leaves your facility. Built for IRB-friendly clinical studies and biomechanics research.
Sports Performance
BiomechanicsBiomechanical analysis for coaching, injury prevention, and athlete development — field or gym.
VFX
Markerless mocapStudio-quality markerless skeletal tracking. Drop into existing pipelines without a hardware overhaul.
Pulled directly from an internal QA capture: four synced cameras tracking a full lift in a real gym, with per-camera reprojection error overlaid live and the triangulated 3D skeleton reconstructed alongside it.
From an internal capture session — swap in your validated benchmark numbers once available.
Bar path & lift tracking
Bar path, bar velocity, range of motion, and rep counting — captured automatically from a multi-camera rig, no markers or sensors on the athlete or bar.
Shot tracking & zone efficiency
Every shot located, scored, and zoned in real time — makes, misses, and efficiency by court location, built from the same camera-and-keypoint pipeline.
A10G cloud processing
Three clicks to a cloud GPU
Point it at a recording session, confirm anchors, hit upload. 3D pose estimation runs on a cloud A10G GPU — no local GPU, no pipeline to babysit.
QA worklist
Fix exactly what the model missed
Frames with high reprojection error are queued automatically. Correct a joint once, across every camera view, and the fix propagates through the clip.
Skeleton overlay
Confirm the fix before moving on
Play the corrected 3D skeleton back over the original footage, frame by frame, to verify tracking is clean before it's used downstream.
↻ Steps 2 and 3 repeat, clip by clip, until reprojection error is low enough to trust
Private
Country-based servers, for example UK-based for UK deployments — compliant with the UK Data Protection Act 2018.
Secure
Cloud video purged weekly — most processing runs offline, on-device.
Easy
Automated QA flags — clean 3D data, no computer vision background required.
Enterprise 3D pose estimation systems cost hundreds of thousands of dollars and take weeks to set up. We built nodal because we've seen firsthand — working at Sony and Google — that the core technology is mature enough to be made accessible. Our mission is to put professional-grade computer vision tools in the hands of every sports team, indie studio, and factory floor.