What is it about?
Have you ever wanted to track your running form on a treadmill without strapping on bulky gadgets like smart insoles or watches? Our work introduces VibRun, a novel system that uses cheap sensors attached to the treadmill to pick up the tiny vibrations from your footsteps. These vibrations tell us key details about how you run, like your step rhythm (cadence), how long your foot touches the ground each time, the timing between strides, where pressure builds up on your foot, and the center point of that pressure. We built this using advanced AI models that process the vibration data in real-time, right on a small device like a Raspberry Pi. No need for expensive equipment or anything touching your body, and it's all unobtrusive and works while you're just running normally. We tested it with 17 people of different ages, sizes, and fitness levels, running at various speeds on different treadmills. The results were spot-on, with tiny errors like just 29 milliseconds off on ground contact time and about 14 millimeters on pressure center. This makes it easy to get personalized tips to improve your run, avoid injuries, or even sync with virtual reality games and fitness apps for a more fun workout.
Featured Image
Photo by Intenza Fitness on Unsplash
Why is it important?
Running is hugely popular, with nearly 50 million people in the US alone hitting the pavement or treadmill each year, and the indoor running market is booming to $4 billion in 2024. But traditional ways to analyze your gait, like wearable sensors or costly lab setups, can be uncomfortable, expensive (over $8,000 for some insoles), or invasive. VibRun changes that by making detailed running insights accessible to everyone using just affordable vibration sensors and AI. It's unique because it's completely hands-off, real-time, and robust against noise, helping everyday runners optimize their form, spot imbalances early to prevent injuries like knee or ankle issues, and boost performance. In a world where virtual sports and gamified fitness apps are exploding, this tech can integrate seamlessly to make workouts more engaging and personalized, like giving your avatar in a running game your exact biomechanics. Timely now with rising interest in home fitness post-pandemic, it democratizes pro-level analysis, potentially reducing healthcare costs from running injuries and empowering more people to stay active safely.
Perspectives
As one of the lead researchers on VibRun, I'm thrilled about how this bridges cutting-edge AI with everyday exercise. Growing up as a casual runner, I could always feel how clunky wearables felt during workouts. This project stemmed from that frustration, aiming to make gait analysis invisible and effortless. What excites me most is its potential beyond labs: imagine casual gym-goers getting instant feedback on their phone to tweak their stride and avoid shin splints, or virtual reality runners feeling more immersed. It's a step toward making health tech truly user-friendly, and I look forward to seeing it evolve into apps that motivate people to run smarter, not harder.
Tianhao Wu
University of Georgia
Read the Original
This page is a summary of: VibRun: Real-time Unobtrusive Gait Analysis for Treadmill Running via Footstep Vibrations, Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies, August 2025, ACM (Association for Computing Machinery),
DOI: 10.1145/3749539.
You can read the full text:
Contributors
The following have contributed to this page







