About the Tonal Research Lab

The Tonal Research Lab works at the intersection of artificial intelligence, computer vision, and biomechanics to advance the science of how people move and perform.

Our Research

Human movement contains an extraordinary amount of information. We study how patterns of motion and performance can reveal technique, capability, fatigue, adaptation, and progress. Our aim is to convert these signals into movement intelligence: a deeper, more useful understanding of the individual that can support smarter training decisions and personalized guidance.

Areas of research include:
Computer Vision: Multi-camera pose fusion, motion and form classification.
Multimodal Sensor Foundation Models: Foundation modeling on motion, pose, and textual data. Fatigue and injury detection.
Coaching and Skill Acquisition: How and when to give feedback for optimal performance improvement.
Biomechanical Models: Force and moment propogation through joints and muscles.

Point cloud
Projected Point Cloud
Multi Angle Pose Fusion

Inside the Lab

Motion capture laboratory with a Tonal trainer at the center of a Vicon camera rig
The motion capture lab at our San Francisco headquarters
Musculoskeletal model with joint coordinate frames standing on a force plate
Musculoskeletal model with joint coordinate frames, over a force plate

A purpose-built motion capture laboratory at our San Francisco headquarters: a Tonal trainer at the center of a twenty-camera Vicon rig, twelve optical cameras tracking motion at 120Hz and eight synchronized video cameras, over force plates that measure every ground reaction. We run marker-based and markerless capture side by side, and it is where our volunteer data collection sessions happen.

The lab is where we produce ground truth. It validates what the trainer's sensors measure, benchmarks our pose estimation, and feeds biomechanical simulations that generate synthetic data for training our models. The fleet gives us scale; the lab gives us truth.

A session starts with calibration. The optical cameras are re-registered to a common coordinate frame and the force plates are checked against a known reference before anyone lifts, so that every stream we record afterward can be trusted and time-aligned to the others. Nothing downstream is better than the calibration it rests on.

During a capture, a volunteer trains on the Tonal while every system records at once: marker trajectories, video from each angle, ground reaction forces, and the trainer's own cable force and position telemetry. Because the streams share a clock, a single repetition can be examined from all of them simultaneously, which is what makes it possible to ask whether two methods agree on the same instant of the same movement.

Running marker-based and markerless capture together is the point of that arrangement. Marker-based capture is the reference standard but requires a lab; markerless capture is what can run in a living room. Measuring one against the other tells us where our pose estimation is already accurate enough to act on and where it still needs work — per joint, per exercise, per camera angle.

From there the data becomes a model. Captured motion drives musculoskeletal simulations that estimate joint angles and internal loads no home sensor can observe directly, and those simulations in turn generate labeled synthetic movement we can train on: body types, tempos, and exercise variations we have not yet collected in person.

We are the research arm of Tonal. Our job is to build the intelligence that helps people get stronger, move better, and live longer, and to push on the open questions in strength physiology that only data at our scale can answer.

Our lab instrument is the Tonal trainer itself: a precision strength machine in hundreds of thousands of homes, measuring cable force and position at 50Hz on every rep, streaming real-time video for pose estimation, and syncing with heart rate monitors and wearables. Together they produce the world's largest strength physiology dataset.

Markerless motion capture in the lab

The Team