Ambulatory walking speed estimation under different step lengths and frequencies

Abstract

In this study we investigated the feasibility and performance of estimating walking speed under different combinations of step length and step frequency using a shank-mounted inertial measurement unit (IMU). The estimation algorithm is based on the fact that the walking is a cyclical motion with a distinguishable pattern and an inverted pendulum-like behavior. To evaluate its performance under different walking conditions, treadmill trials were conducted with controlled step lengths and step frequencies. A root mean squared error (RMSE) of 5.6% was achieved in the walking speed estimation across all combinations of step length and step frequency. As the shank-mounted IMU walking speed estimation method showed a robust performance at wide range of step lengths and step frequencies, it could potentially be used as a low-cost alternative for walking speed measurement in a non-laboratory environment.