Turn the pitch into a biomechanics lab
FitWise LAB monitors technique deviation from broadcast video
FitWise turns video into 3D biomechanics. No sensors. No wearables.
The precision of a motion lab, on the pitch.
Track every athlete against their healthy baseline
FitWise uses match and training footage to rebuild every athlete in 3D. From these 3D models, we measure biomechanics for every step: sprint phases, joint angles, speed, and asymmetry. The same measurements captured in a motion lab.
Our methods are grounded in sports science, incorporating established assessment scores such as S-MAS, with accuracy validated against motion capture.
But measuring biomechanics is only the beginning.
Every athlete has a unique running style. Rather than comparing athletes against population norms, FitWise builds a personalized baseline from each athlete's own healthy games. We then track how their movement changes from game to game, spotting deviations from their usual running form.
Practitioners can see exactly which movement patterns have changed—and by how much.
What changed in this game?
After every game, FitWise LAB shows what changed in each athlete's running form. It highlights the biggest differences from their healthy baseline.
Are they back to their pre-injury mechanics?
FitWise LAB tracks an athlete's biomechanics throughout recovery, quantifying how their movement differs from their healthy, pre-injury baseline.
This gives performance and medical teams objective measurements to assess recovery and make informed return-to-play (RTP) decisions.
Automatic sprint mechanics assessment score
S-MAS is the state-of-the-art score of sprint technique (Bramah et al., 2024). We automatically compute it on every stride of every game or training session.
Do they share movement patterns with athletes who were injured?
FitWise PRISM is an ongoing study of 100 hamstring injuries in elite football.
We are exploring movement patterns before injury and their potential links to hamstring injuries. As the findings emerge, FitWise LAB will enable practitioners to compare athletes with the PRISM cohort and identify shared patterns.
Advised by practitioners and researchers
Robin Thorpe PhD
Ten years at Manchester United, then Director of Performance at Red Bull. A practitioner and researcher specialising in sports performance, athlete monitoring, human movement and applied sports science. Robin provides scientific leadership and strategic guidance to FitWise and co-authors the PRISM hamstring study.
Michael Clarkson
Vice President at Bloom Sports Partners; formerly Catapult and STATS Perform. Michael has decades of experience advising high-performance organisations and understands deeply how clubs and leagues leverage performance technologies.
How we get there
Step by step, tracking a single athlete across four games.
From broadcast video to a 3D body
FitWise takes existing footage, finds every sprint and high-speed run, and reconstructs the athlete's body in 3D. From this 3D reconstruction, we extract speed, acceleration, and a comprehensive range of biomechanical parameters.
The only input we need is video. No additional sensors. No hardware installation. No wearables.
FitWise works with existing broadcast and tactical footage, as well as video from training sessions, bringing detailed biomechanical analysis directly to the environments where athletes perform.
Automatic sprint kinograms
FitWise automatically detects five key phases of every sprint stride: toe-off, max vertical projection (MVP), late swing, touchdown, and mid-stance.
For each phase, we capture the original video frame and the corresponding 3D body.
Together, these frames form a sprint kinogram, making it easier to assess an athlete's running form.
Making every step measurable
FitWise turns every step into measurable biomechanics. From speed to joint angles and asymmetry, practitioners can track an athlete's movement in every game and training session.
FitWise optical vs. inertial skeleton tracking
A Positional Agreement Study for Sports Biomechanics
The full process and results: how the skeletons were synchronised and aligned, the per-joint results for all three activities, the reference-side effects we found in the suit's export, and what this study can and cannot claim.
From population-level to personalized monitoring
Every athlete has a unique running form, shaped by their body, training, and injury history. To reflect this, FitWise builds a personal baseline from each athlete's healthy games. We then compare every new game against it to spot changes in their running form.
Each athlete is compared with their own healthy movement, not a population norm.
In this visualization, the transparent bodies show sprint strides from healthy games, while the solid body represents their average.
Tracking the movement changes from game to game
FitWise tracks how each biomechanical metric drifts from the athlete's personal baseline, game after game. We rank these changes and highlight the biggest flags in the squad dashboard, giving practitioners a clear view of what needs attention.
Sports scientists can access and export the complete raw measurements to build their own models.
Try it on your own athlete
Pick one athlete. Send us video of their recent games, or of a sprint-testing session, and we will build their analysis and in-depth report for free.
All we need is video: no sensors, no medical data. We are offering this to 10 clubs in 2026.
FAQ
What does it need from a club?+
Video. Match broadcast, a tactical camera, or a fixed practice camera with the pitch markings in view. No wearables, no cameras to install, nothing for the players to do.
How accurate is it?+
2–5 cm per joint against an Xsens motion-capture suit, across nine athletes and 160,740 frames. See the whitepaper.
Which sports does FitWise LAB support?+
Football and American football.
Is this injury prediction?+
No. FitWise reports that mechanics changed, how much and where. It does not put a risk percentage on a player or recommend clinical or selection decisions; the practitioner decides. Where the science is missing, we run the study instead.
Does it work on practice footage?+
Yes, through the same process, as long as the players are on a pitch with visible markings and remain distinguishable at the far end of the frame.
How are players identified?+
By jersey number and roster.
Keeping players available is worth more every season
Performance and medical teams work every day to keep players on the pitch. But the demands keep growing: more matches, more high-speed running, and less time to recover.
the wage cost of 4,456 injuries across Europe's big five leagues in 2024/25
Howden Injury Index 2024/25 ↗more sprint distance per team in a Premier League match than 10 years ago
Allen et al., BJSM 2025, 2015/16 vs 2024/25 ↗of Euro 2024 players had the recommended 28 days of offseason rest before the 2024/25 season
FIFPRO Player Workload Report 2025 ↗of injuries in men's professional football are hamstring injuries, 2x vs 20 years ago
UEFA Elite Club Injury Study, BJSM 2023 ↗GPS and wearables already tell staff how much each athlete runs. FitWise adds how they run—using video the club already has.
With FitWise, practitioners can see when an athlete's running mechanics drift from their own healthy baseline.
We are building the movement data layer for elite sport
We built OpenCV, the world's most used computer-vision library: 50M+ downloads a month, 3B+ devices, NASA's Mars rovers. FitWise points this experience at one problem: measuring every player's movement without asking anyone to wear anything.