Project 02 · 3 publications

Self-Powered Triboelectric Motion Sensing

Batteryless wearable systems that exploit the triboelectric mechanism simultaneously for sensing and energy harvesting — enabling human activity recognition, gait analysis, and rehabilitation monitoring without any external power supply.

Self-Powered SensingActivity RecognitionGait AnalysisWearable IoTRehabilitationkNN / Machine Learning
Overview

One Device, Two Functions: Sensing and Harvesting

Conventional wearable sensors — accelerometers, pressure sensors, EMG electrodes — require a battery or continuous power source. The triboelectric mechanism offers an elegant alternative: the same charge generation that produces harvestable energy also encodes a signal proportional to the motion that caused it. This enables batteryless wearable IoT nodes where a single triboelectric element serves as both sensor and power source.

The high output voltage (>20 V peak-to-peak) at low human-motion frequencies (1–10 Hz) makes triboelectric devices especially suited for this application — the signal is measurable without amplification in many cases, reducing system complexity further.

Dual-function principle: The triboelectric structure generates a self-powered motion signal from physical activities and simultaneously harvests energy, providing sustainable power for wireless data transmission in the wearable IoT node — advancing maintenance-free, autonomous health monitoring systems with zero battery replacements.

>80%Activity recognition accuracy
20 V+Peak signal output
5Classified activities
Application Areas

Sensing Across Health Contexts

The self-powered sensing paradigm was validated across two complementary deployment scenarios:

Human Activity Recognition

Body-worn triboelectric sensors classify five daily activities — sitting/standing, walking, stair climbing up/down, and running — using kNN clustering on signal features, with no external power supply or signal conditioning circuits.

Gait Monitoring & Rehabilitation

Four triboelectric elements embedded in shoe soles capture ground contact force (GCF) patterns and gait timing parameters, enabling clinical-grade gait analysis for rehabilitation patients during daily life.

Concurrent Energy Harvesting

The same tribo-elements output both a sensing signal and harvestable charge. TriboWalk generates >20 V during walking, enabling the sensor node to be partially or fully self-powered during operation.

Wireless Transmission

Motion data is collected by a 10-bit ADC and transmitted wirelessly. The system architecture exploits the self-powered capability to minimize total node power consumption and extend deployment lifetime.

Publications
IEEE Internet of Things Journal · Vol. 5, No. 6 · December 2018
TriboMotion: A Self-Powered Triboelectric Motion Sensor in Wearable Internet of Things for Human Activity Recognition and Energy Harvesting
Hui Huang, Xian Li, Si Liu, Shuai Hu, Ye Sun
Develops a self-powered motion sensor system for wearable IoT that eliminates signal conditioning circuits by exploiting the triboelectric physical model directly. Classifies five daily activities with >80% recognition accuracy for walking and sitting/standing, while simultaneously harvesting energy from the same motion signal. Establishes the foundations for next-generation self-powered wearable IoT.
IEEE EMBC · August 2016
TriboWalk: Triboelectric Dual Functional Wireless System for Gait Monitoring and Energy Harvesting
Xian Li, Hui Huang, Ye Sun
Reports the design and validation of TriboWalk, a shoe-sole triboelectric wireless system for gait monitoring and energy harvesting. Four removable tribo-elements per sole capture ground contact force patterns and gait timing parameters. Generates >20 V during walking for concurrent energy harvesting. Validated at normal and fast walking speeds; includes a visualization tool for gait data analysis.
IEEE EMBC · August 2016
A Triboelectric Motion Sensor in Wearable Body Sensor Network for Human Activity Recognition
Hui Huang, Xian Li, Ye Sun
Early work establishing the triboelectric motion sensing paradigm. Designs a wearable triboelectric sensor node that collects motion signals from physical activities without any power supply, exploiting the high (~20 V) peak-to-peak output at low frequencies. Uses kNN clustering to recognize five common activities, achieving >80% success rate for walking and sitting/standing. Demonstrates dual function as an energy harvester.