Cadence and Stride Length Measurement Using a Foot-Mounted IMU

Abstract

The purpose of this project was to develop a prototype for measuring cadence (step frequency) and stride length directly from the foot, enabling the collected data to be visualized and presented to the user in a mobile application. These metrics are relevant because they can provide runners with insight into their running technique and help users identify patterns that may contribute to more efficient training. Two different prototypes were developed and evaluated. The first prototype was a lace-mounted design using an Inertial Measurement Unit (IMU)-based step detection algorithm. The second prototype was a sole-based prototype, where the IMU was placed in a cutout on the insole and pressure sensors were used for step detection. For stride length estimation, both systems employed an Extended Kalman Filter (EKF) on the IMU data. The systems communicated with a mobile application via Bluetooth Low Energy (BLE), enabling the collected data to be presented to the user in a graphical interface. The results showed that both designs performed similarly in terms of step detection, achieving more accurate measurements during running and jogging compared to walking. However, regarding stride length estimation, the sole-based prototype outperformed the lace-mounted prototype across all tests, achieving Mean Absolute Relative Error (MARE) values between 5.57% and 15.15%, compared to 27.57% to 35.32%. Overall, the results indicated that the sole-based prototype provides more reliable performance than the lace-mounted prototype. However, this came with a trade-off between usability and performance, which was a recurring challenge throughout the project. The developed prototypes demonstrated the potential of foot-mounted sensing as an alternative to smartwatch-based measurements and highlighted promising opportunities for future development in running analysis applications.