Walking pattern analysis using deep learning for energy harvesting smart shoes with IoT

Abstract

Wearable Health Devices (WHDs) benefit people to monitor their health status and have become a necessity in todays world. The smart shoe is the type of WHD, that provides comfort, convenience, and fitness tracking. Hence smart shoes can be considered as one of the most useful innovations in the field of wearable devices. In this paper, we propose a unique system, in which the smart shoes are capable of energy harvesting when the user is walking, running, dancing, or carrying out any other similar activities. This generated power can be used to charge portable devices (like mobile) and to light up the LED torch. It also has Wi-Fi-that allows it to get connected to smartphones or any device on a cloud. The recorded data was used to determine the walking pattern of the user (gait analysis) using deep learning. The overall classification accuracy obtained with proposed smart shoes could reach up to 96.2 %. This gait analysis can be further used for detecting any injury or disorder that the shoe user is suffering from. One more unique feature of the proposed smart shoe is its capability of adjusting the size by using inflatable technology as per the users comfort.