Multisensor Fusion for Noninvasive Worker Health Monitoring in Industry 4.0

Workers’ health state monitoring in Industry 4.0 is crucial for the prevention of work-related musculoskeletal disorders caused by repetitive movements. Electromyographic signals (EMG) are commonly used for detecting muscle activity and assessing muscle load. Unfortunately, accurate processing and analysis of EMG remain challenging due to the nonlinear nature of this signal. To enhance the accuracy of muscle activity analysis, this study employs wearable sensors to concurrently monitor electrocardiographic (ECG), EMG, accelerometric, gyroscopic, and magnetometric signals using two synchronized Shimmer3 ECG/EMG units. These signals were recorded and analysed on 13 volunteers during 8 different scenarios of repetitive upper limb movements commonly encountered in Industry 4.0 environments. Heart activity was evaluated using ECG, achieving an average accuracy of 88.8% in R-peak detection, with average values of μ and ±1.96σ being 0.2 bpm and 12.6 bpm, respectively, in heart rate determination. Heart rate variability analysis was conducted, revealing varying degrees of stress or fatigue in most cases. Physical activity detection was performed using two separate approaches: 1) an EMG signal analysis method and 2) a newly introduced accelerometer-gyroscope-magnetometer (AGM) method. The AGM method developed in this study demonstrated superior performance compared to the EMG method, with an average accuracy of 96.4% compared to 45.5% for EMG. Muscle fatigue quantification was effectively performed by analyzing the frequency characteristics of EMG. The results of this study confirmed the potential of wearable sensors for monitoring the health status of workers in Industry 4.0 environments, with a focus on the application of the novel AGM method.