A walking monitoring shoe system for simultaneous plantar-force measurement and gait-phase detection

This paper presents a walking-monitoring-shoe (WMS) system capable of simultaneous performing accurate plantar-force measurement and reliable gait-phase detection for continuous monitoring of human walking on treadmill. Based on anatomical information, the WMS employs four strain-gauges embedded in a homemade sole to accurately measure the contact force of the human foot exerted on the shoe-pad, and an efficient classification algorithm to detect five distinct gait-phases from the measured plantar force patterns. The WMS was experimentally evaluated, which has a typical 2^nd-order system dynamics (with an 8% overshoot and a 2% settling time of 76ms when subject to a step input). Experimental results show that the accuracy and resolution of the sensing system are 1.09±0.09% (significant level=0.95) and 2N (0.2% of the maximal value of the load), respectively. The root-mean-square (rms) difference between the output signals of the WMS and the calibrated dynamic loading system was 1.67±0.12% (significant level = 0.95). The feasibility of this integrated sensing/detection system was experimentally validated against video data, which relates gait-phases to the leg kinematics.