Rutgers Built A Shoe That Watches You Walk. No Battery Required. The Gait Data Is The Point, Obviously.
Engineers at Rutgers built a battery-free smart shoe prototype that tracks gait, meaning walking pattern, including balance, speed, stride and rhythm. The system uses AI to achieve 95.4% accuracy in movement classification. Niu, doctoral student Fuying Dong, and Chi Han are listed as inventors on a Rutgers provisional patent application for the technology.
The principle here is ambient sensing. You do not need a battery or a conscious user to collect clinically useful signal. The shoe harvests energy from walking and feeds AI models that detect gait changes, which can indicate disease progression, fall risk, or rehabilitation status. That matters because clinicians currently observe gait in short lab visits. Continuous monitoring during daily life is the mechanism that transforms episodic observation into longitudinal data.
Rutgers engineers Niu, Fuying Dong, and Chi Han developed the prototype and filed a provisional patent. The work targets clinical gait analysis currently done through observation in clinics or laboratories.
- Open a notes app on your phone and walk normally for two minutes while recording your steps using any free pedometer app. Note your step count and pace.
- Walk the same distance again, this time deliberately altering your stride length or speed. Record the new numbers.
- Compare the two datasets. The difference in counts and timing is, in crude form, what gait analysis measures. The Rutgers shoe does this continuously and with 95.4% accuracy, without you thinking about it.