Machine learning enabled smart human activity monitoring using corn husk based eco-friendly triboelectric nanogenerator

The keys to success in health status clinical investigations, fitness services, and several other related fields are accurate identification of human actions. In this study, the machine learning is integrated with corn husk based biodegradable and environment friendly triboelectric nanogenerator (TENG). This self-powered sensor designed and fabricated in this present study is used for smart human activity monitoring. We devote our focus on using biodegradable materials to make environmentally friendly devices as a green energy source. The corn husk used in this study is a natural agriculture waste by-product. This corn husk based sensor is capable of generating 140 volts and charging various commercial capacitors. This eco-friendly sensing device is also tested as a wearable sensor to gather data and track the user's physical activity, such as jumping, running, and walking. For accurate prediction of human activity, the designed corn husk based sensor is coupled with an extra randomized tree and random forest-based machine learning model. These models achieve a 98.89% and 98.33% accuracy rate in identifying the user's three activities. The research presents a novel strategy to support tailored applications of TENG sensors in human motion tracking and enables the development of intelligent, self-sufficient systems for novel uses.