You Can Wash Hands Better: Accurate Daily Handwashing Assessment with a Smartwatch

2021-12-09

AI Summary

This study addresses the lack of fine-grained, automated assessment of routine handwashing behavior. We propose a real-time hand hygiene quality evaluation method leveraging smartwatch inertial sensors. Handwashing is formulated as an action segmentation task, integrating synchronized accelerometer and gyroscope signals via a dual-stream UNet-like architecture augmented with lightweight temporal modeling. The framework enables precise hand gesture sequence recognition, accurate onset/offset detection (timing error < 0.5 s), and automatic quality scoring (score error < 5 points). To our knowledge, this is the first work to quantify handwashing quality through temporal gesture analysis, achieving cross-user and cross-scenario robustness while supporting actionable real-time feedback. Evaluated in long-term deployments across clinical and field settings, the system demonstrates stable operation with a mean recognition accuracy of 92.27%.

Abstract

Hand hygiene is among the most effective daily practices for preventing infectious diseases such as influenza, malaria, and skin infections. While professional guidelines emphasize proper handwashing to reduce the risk of viral infections, surveys reveal that adherence to these recommendations remains low. To address this gap, we propose UWash, a wearable solution leveraging smartwatches to evaluate handwashing procedures, aiming to raise awareness and cultivate high-quality handwashing habits. We frame the task of handwashing assessment as an action segmentation problem, similar to those in computer vision, and introduce a simple yet efficient two-stream UNet-like network to achieve this goal. Experiments involving 51 subjects demonstrate that UWash achieves 92.27% accuracy in handwashing gesture recognition, an error of<0.5 seconds in onset/offset detection, and an error of<5 points in gesture scoring under user-dependent settings. The system also performs robustly in user-independent and user-independent-location-independent evaluations. Remarkably, UWash maintains high performance in real-world tests, including evaluations with 10 random passersby at a hospital 9 months later and 10 passersby in an in-the-wild test conducted 2 years later. UWash is the first system to score handwashing quality based on gesture sequences, offering actionable guidance for improving daily hand hygiene. The code and dataset are publicly available at https://github.com/aiotgroup/UWash

Problem

• Assessing handwashing quality using smartwatch data
• Improving adherence to hand hygiene guidelines
• Recognizing handwashing gestures with high accuracy

Innovation

• Smartwatch-based handwashing assessment system
• Two-stream UNet-like network for segmentation
• High accuracy gesture recognition and scoring