Listen to Your Footsteps: Wearable Device for Measuring Walking Quality

Abstract

In this paper, we present a low-cost context-aware technique for determining a user's walking quality. This is achieved by filtering and analyzing the acoustic signal generated when users walk. To extract the acoustic values of footsteps, we implemented a simple wearable device attached on the user's ankle. To verify our approach, we conducted a preliminary test using several pattern classification algorithms. The results show that our system achieves an 89.6% average for three different walking styles (best, good, and bad) and 86.9% for four different real-world ground sets (carpet, asphalt, sand, and wood). We believe that our technique can be applied to existing context-aware techniques as well as various unexplored domains in wearable devices.