Dr Dinghuang Zhang from the University of Portsmouth describes how his research into three-dimensional ground reaction forces is already rewriting the rules.
The ability to accurately measure how someone moves has been something of a holy grail in sport and healthcare. Historically, the technology just hasn't been up to the task. That could be about to change.
Dr Dinghuang Zhang, knowledge transfer programme (KTP) associate at TG0 and the University of Portsmouth, has been developing a way to measure three-dimensional ground reaction forces (GRFs) using TG0's smart insole and embedded artificial intelligence. The research is funded by Innovate UK.
Just six months into the 2.5-year project, the team is already achieving 95% accuracy in estimating 3-dimensional GRFs. *"The advantage of TG0's insole is it provides us with pressure data with higher resolution of the whole foot, and the direction and acceleration of the foot when moving,"* Zhang says.
Why This Matters
Historically, sport scientists and healthcare professionals needed a force plate or treadmill to measure GRF — both highly inflexible, expensive, and disruptive to natural gait patterns. *"If our insole can provide a similar result and can be embedded into any shoe, it's a huge revolution for the industry."*
Working with the University of Portsmouth, Zhang has been able to collect and synchronise data between TG0's smart insole and the university's force plate — the first time two inputs have been used simultaneously to capture detailed pressure distributions across the foot and dynamic movement parameters.
The Research Method
Study participants were asked to perform a group of movements including jumping, running in place, walking in place, swaying, and squatting, with inputs collected from both the insole and the force plate. Once the AI model has been trained to map the relationship between the two, the force plate becomes unnecessary.
*"At this stage, we are only focusing on predicting the 3-D GRF between a foot and the ground. We are also planning to include foot movements from sports such as baseball or golf to enrich our dataset. If we can achieve higher accuracy, it will surpass anything else on the market."*
Future Potential
Zhang believes the research opens up wider possibilities for embedded AI in healthcare. *"Once we have the framework and platform, we may be able to choose another device or scenario. In the future, we hope to select a foot condition measurable via the 3-D GRF or gait, so the smart insole can be used as a medical aid."*
Crucially, all data processing happens on small modules embedded within the device — meaning data doesn't need to be stored or uploaded to the cloud, which is good for both privacy and accessibility.
The research is still in its early stages but its potential to advance wearable tech, personalised training, and biomechanics is extensive.
