A Self-Powered and Self-Sensing Lower-Limb System for Smart Healthcare

In the age of the artificial intelligence of things (AIoT), wearable devices have been extensively developed for smart healthcare. This paper proposes a self-powered and self-sensing lower-limb system (SS-LS) with negative energy harvesting and motion capture for smart healthcare. The SS-LS achieves self-sustainability via a half-wave electromagnetic generator (HW-EMG) that recovers negative work from walking with a low cost of harvesting. Additionally, the motion capture function of the system is achieved by the three-channel triboelectric nanogenerator (TC-TENG) based on binary code, which can accurately detect the angle and direction of the knee joint rotation. The bench test experiment indicates that the HW-EMG has an average output power of 11.2 mW, sufficient to power a wireless sensor. The three-channel voltage signal of TC-TENG fits well with the binary signal, which can precisely detect the angle and direction of rotation. Furthermore, the SS-LS demonstrates an identification accuracy of 99.68% and a motion detection accuracy of 99.96% based on an LSTM deep learning model. Demonstrations of Parkinson's disease and fall detection and monitoring of three training modes (sit-and-stand, balance, and walking training) are also performed, which exhibit outstanding sensing capabilities. The SS-LS is highly promising in sports rehabilitation medicine and can contribute to the development of smart healthcare.