Introducing a modular activity monitoring system

Area of Science:

• Biomedical Engineering
• Wearable Technology
• Human Activity Recognition

Background:

• Activity monitoring systems are crucial for health and performance assessment.
• Existing systems often lack modularity and comprehensive data integration.
• Integrating diverse sensor data, like accelerometry and heart rate, can enhance accuracy.

Purpose of the Study:

• To introduce a modular activity monitoring system with adaptable functionality.
• To evaluate the system's performance in physical activity intensity estimation and activity recognition.
• To analyze the contribution of different sensor placements and heart rate data.

Main Methods:

• Development of a modular system with three distinct functional modules.
• Utilized 3D-accelerometers on the lower arm, chest, and foot, along with a heart rate monitor.
• Collected data from 8 subjects performing 14 different activities for system evaluation.

Main Results:

• The system achieved 94.37% accuracy for intensity estimation using chest accelerometer and heart rate monitor.
• Activity recognition accuracy reached 90.65% when using all three accelerometer placements and the heart rate monitor.
• Analysis highlighted the significant impact of accelerometer placement and heart rate data on performance.

Conclusions:

• The modular system effectively estimates physical activity intensity and recognizes basic activities.
• Optimal sensor placement and the inclusion of heart rate data are critical for high-performance activity monitoring.
• The system's modularity allows for flexible adaptation to various monitoring needs.