Real-time mobility aware shoe — Analyzing dynamics of pressure variations at important foot points

Abstract

Data collected from pressure sensors attached to shoe insole is a rich source of information about the dynamics of the varying pressure exerted at different points while a person is in motion. Depending on the accuracy and the density of the points of data collection, this could be applied for different uses. Analyzing the time series data of the pressure, it is possible (1) to detect faults in walking and balancing problems for old people, (2) to design personalized foot orthoses, (3) to calculate the calorie burnt, even when walking and jogging are mixed, and the road slope changes, (4) to find subtle faults in sprinters or tennis players, (5) for person identification, (6) even for initiating alarm arising from mishandling of machines (like accelerator pedal of a car). In this work, we look for an efficient, real-time, yet cheap solution. We use a few thin, cheap, resistive pressure sensors, placed at critical points on the insole of the shoe to collect dynamic pressure data, preprocess it and extract features to identify the mobility speed. Nearly 100% classification accuracy was achieved. Thus, the target to classify whether the person is walking or jogging or climbing up or down the stairs was found to be possible, even with very simple gadget. From the time duration and the speed, the distance travel could be calculated. If, in addition, this signal could tell us the body-weight, we could accurately calculate the calorie burnt at the end of the day. The analysis method, and results from real experiments are discussed.