University: The University of Alabama
Team Members: Nagaraj Hegde, Matthew Bries, Adviser: Dr Edward Sazonov
TI Parts Used:
We are presenting an insole based activity monitor – SmartStep. The goal was to make SmartStep a completely wireless (wireless data retrieval, wireless charging and wireless firmware upgrade) wearable system for activity monitoring and develop an Android Application for the same. The SmartStep captures the pressure sensors, accelerometer and gyroscope data, while the user is wearing the insoles in their shoes and transfers the collected data to the Android smartphone. The associated Android smartphone collects the data in real time and performs activity classification utilizing the machine learning models implemented in it.
The SmartStep solves the existing problems in the footwear based wearable systems such as battery recharging problem. With the associated base station platform recharging of the SmartStep’s battery is as seamless as placing the shoes on the shoe rack. With wireless firmware upgrade feature, the same SmartStep can be utilized in variety of applications or the sensor modalities in the SmartStep can be altered while in use. With low power implementation, the SmartStep can be used for ~5 days on a single charge (12 hours of wear per day).
When it comes to activity classification, SmartStep has very high accuracy of more than 95% in classifying common daily living activities of sitting, standing, walking and cycling. Specially it has more than 99% accuracy in classifying sitting and standing activities, which are difficult for many commercially available activity trackers to distinguish.
Many of the published works on SmartStep can be found in https://www.researchgate.net/profile/Nagaraj_Hegde2
- Wireless Data Retrieval (over Bluetooth Low Energy)
- Seamless wireless charging
- Wireless firmware upgrade
- Embeds pressure sensors, accelerometer, gyroscope, BLE, rechargeable battery and wireless charging all in an insole
- More than 95% accurate in classifying daily living activities of sitting, standing, walking and cycling
- Low power implementation
Additional Papers are located here
The user has to just insert the insoles in their shoes and connect the associated Android smartphone application to the particular SmartStep. The Android application asks for the user’s profile (including age, gender, weight, height etc) which will be used in calculation of daily energy expenditure goal. The SmartStep wirelessly transmits the sensor signals from the insole to the Android smartphone. The smartphone application classifies the user’s activity into sitting/standing/walking/cycling in real time, and computes the estimated energy expenditure in real time and gives visual feedback to the user. The application also tracks the progress towards the daily goal and advices the user about the progress. To recharge the insole, all that the user needs to do, is to place the shoes containing SmartSteps on the base station, which will wirelessly charge the SmartSteps.