Because factory work can be physically strenuous, potentially leading to overworked and fatigued workers who are more prone to injury and accident risks, as well as chronic health problems and impaired performance, the researchers developed a network that measures heart rate, heart rate variability, skin temperature and locomotor patterns from six regions on the torso and arms.

Source: Payal Mohapatra and Vasudev Aravind

Explaining that there are no widely accepted biomarkers or metrics for fatigue, the authors calibrated their measurements to self-reported perceived exertion, on a 0 to 10 scale.

During trials of the system, 43 participants, ages 18 years old to 56 years old, performed two manufacturing tasks — composite sheet layup and wire harnessing — while outfitted with weighted vests to mimic levels of fatigue that might be felt at the end of a full shift. During those trials, the participants reported fatigue levels at various time points throughout an approximately hour-long data collection period.

The team then employed a machine learning model to that data from the participants to predict fatigue levels in real time.

The researchers suggest that the technology might improve factory safety, mitigate risks and empower workers.

The team’s findings are detailed in the article “Wearable network for multilevel physical fatigue prediction in manufacturing workers,” which appears in the journal PNAS Nexus.