Smart-walk: An intelligent physiological monitoring system for smart families

A healthy lifestyle can be maintained by analyzing the physical activities of individuals. Understanding the lifestyle of a family can help in improving quality of life. Wearable devices for activity tracking have become a rapidly growing industrial sector. However, with multiple functionalities embedded in a small device, their accuracy and power consumption can be suboptimal. This paper proposes a piezo-electric based accelerometer sensor design which helps in tracking the physical activities of family and friends. The accuracy of the activity-sensing algorithm is analyzed by various parameters analyzed through the sensor output. The proposed framework was validated using the TI MSP432, Educational BoosterPack and MATLAB® and the learning parameters were modeled using WEKA. The feature based human activity monitoring algorithm gives 97.9% efficiency in the worst case scenario.