Classifying Diverse Physical Activities Using "Smart Garments".

Area of Science:

• Biomedical Engineering
• Wearable Technology
• Health Monitoring

Background:

• Physical activity is crucial for health promotion, reducing disease risk and healthcare costs.
• Accurate classification of physical activities is essential for monitoring and promoting healthy lifestyles.
• Smart textile systems (STS) offer a promising, low-cost approach to activity monitoring.

Purpose of the Study:

• To evaluate the feasibility and accuracy of a novel smart textile system (STS) for classifying physical activities.
• To assess the performance of an STS featuring a smart undershirt (SUS) and smart socks (SSs).

Main Methods:

• Eleven participants engaged in lab-based experiments involving basic postures and varied walking/running speeds.
• Three classification algorithms (K-nearest neighbor, linear discriminant analysis, artificial neural network) were trained.
• Data was analyzed from smart garments used individually and in combination.

Main Results:

• The smart textile system achieved an overall classification accuracy of approximately 98%.
• The STS effectively discriminated between various postures and activities, including walking and running at different speeds.
• High accuracy was observed using data from smart garments separately and combined.

Conclusions:

• Smart textile systems demonstrate high effectiveness in discriminating diverse physical activities.
• Smart garments represent a promising research area and a viable alternative for activity classification.
• This technology holds significant potential for advancing health promotion initiatives.