Built on 20+ years of biomechanics expertise and machine learning research from the University of Bath — bringing cutting-edge academic technology to the field.
Forceteck combines unparalleled datasets, deep biomechanics expertise, physics-informed AI, and generative simulation into one unified platform.
Unique access to gold-standard, lab-grade technology and high-quality datasets, including motion and force data from rugby tackling, scrummaging, sprinting, jumping, and change of direction. Our training data is the foundation of accuracy others can't match.
A dedicated team of experts with experience across all areas of biomechanics — biomechanical modelling, musculoskeletal simulation, computer vision, and applied machine learning. Not just software engineers: scientists who understand movement.
Forget black-box AI models. Our machine learning is informed by the laws of physics — producing predictions that are more robust, easier to understand, and immediately interpretable by coaches and performance staff.
Because our AI models are driven by the laws of physics, we leverage ground-breaking generative approaches to create realistic, predictive simulations of sporting actions — letting you model performance and injury risk before a single rep is performed.
Every model Forceteck deploys has been developed and validated through peer-reviewed research at the University of Bath — benchmarked against gold-standard laboratory equipment, not just tested in the field.
Dr Dario Cazzola joined the University of Bath as a researcher in computational biomechanics, focusing on injury prevention in rugby. He became a lead researcher on the Rugby Scrum Project — funded by World Rugby (then the International Rugby Board) — a landmark study that produced biomechanical evidence that directly changed scrum laws worldwide.
The research expanded into personalised musculoskeletal models — physics-based digital representations of the human body built from MRI data — used to estimate internal forces during high-impact events like rugby tackles and contested scrums. This pipeline laid the foundation for understanding not just what athletes do, but the forces their bodies actually absorb.
A wave of computer vision breakthroughs enabled the extraction of full-body movement data from standard video footage — no markers, no lab. This opened a new research direction: combining markerless pose estimation with the existing biomechanics models to bring lab-grade analysis out of the lab and onto the pitch.
PhD researcher Andrea Braschi validated the first model capable of estimating rugby tackle forces directly from video — using physics-informed machine learning to constrain AI predictions with the laws of biomechanics. The approach was extended to sprinting, acceleration, deceleration, and change of direction. This body of peer-reviewed work became the scientific core of Forceteck, spun out of the University of Bath to bring it to sport.
Behind every Forceteck model is 15 years of laboratory data collection — thousands of trials where movement and forces were captured simultaneously using gold-standard motion capture and force measurement equipment. This unique synchronised dataset is what allows our machine learning models to learn the relationship between how an athlete moves and the forces their body generates — so that today, forces can be extracted from video alone, with no sensors required.
Musculoskeletal models are physics-based representations of the human body that go beyond tracking movement — they estimate the forces generated at every joint, and can even break these down to the contribution of individual muscles or muscle groups. This level of detail has historically been confined to clinical research labs.
Forceteck's team brings decades of experience in computational modelling. Dr James Cowburn and Dr Josh Carter have each completed PhDs that combine musculoskeletal modelling with machine learning — integrating two disciplines to make these models faster, more scalable, and deployable from ordinary video.
Over the last five years, Forceteck has developed cutting-edge machine learning models capable of estimating the forces generated by athletes across a wide range of movements — from rugby tackling and contested scrums to acceleration, deceleration, and change of direction during locomotion tasks.
This work, led by Dr Andrea Braschi, has been validated against the 15-year synchronised dataset described above — making it one of the most rigorously tested approaches in applied sports biomechanics. The result is groundbreaking: for the first time, it is possible to move away from the lab entirely and collect the same quality of force data on the pitch, from video alone. No sensors. No force plates. No compromise on accuracy.
For over a decade, Dr Dario Cazzola has led research into rugby injury prevention, building unique musculoskeletal and contact models to understand the forces travelling through the body during high-impact events. This work produced one of the most comprehensive contact-sport biomechanics datasets in existence — capturing not only ground reaction forces, but also impact forces at the shoulders of players during machine scrummaging and live rugby tackling.
This unique dataset became the foundation for all of the force estimation technology that has since been developed and commercialised within Forceteck. We are grateful to the RFU Injured Players Foundation for their funding and support throughout the years, without which this research — and ultimately this technology — would not have been possible.
Forceteck's technology is built off extensive biomechanics and computer science research conducted at the University of Bath. Every model and metric is validated against gold-standard lab data — not just tested in the field.
View Our Research PapersTalk to our team and see how Forceteck works on your footage.