As good as these tools are, however, they are not great at assessing physical fatigue or predicting energy expenditure. They don’t account for the influence of external environmental factors or differences between individual workers.

To measure cerebral oxygen levels, the researchers used wireless blood oxygen sensors attached to a headband. A motion capture system collected data for the ergonomic indexes. Photo courtesy Beijing Institute of Technology

The car seat assembly process involves lifting, carrying and static assembly. Photo courtesy Beijing Institute of Technology

While wearing augmented reality goggles, workers had to pick up a weight, move it 3 meters, and “install” it in a virtual vehicle body. Photo courtesy Beijing Institute of Technology

We wanted to develop a more accurate and objective method of assessing fatigue by measuring oxygen saturation levels in the brain. Although the brain represents just 2 percent of body weight, it consumes 20 percent of the body’s energy. Rapid consumption of energy by the brain during exercise can lead to changes in cerebral oxygen saturation. As a result, oxygen levels in the brain can be used to evaluate physical fatigue.

To test our idea, we conducted a field investigation at a Chinese automaker. Specifically, we looked at three actions on the line: lifting, carrying, and static assembly. The first two are self-explanatory. A worker picks up a part or subassembly—a door, a wheel, a seat—in one spot and carries it to another. In static assembly, the worker stays in one spot performing various tasks. He may or may not need to bend at the waist.

We hoped to answer three questions:

•   Can oxygen saturation in the brain assess the degree of fatigue in the human body?
•   Do different weights and carrying heights affect the degree of fatigue?
•   Could our system be more accurate than Tecnomatix Jack at measuring fatigue and energy expenditure?

Experimental Design

To assess the effectiveness of our system, we needed way to quantify physical exertion. For lifting, we used the Lifting Index from the National Institute for Occupational Safety and Health (NIOSH). For carrying, we used the Ovako Working Posture Analysis System (OWAS), which was developed by Ovako OY, a European steel producer. For static assembly, we used the Rapid Upper Limb Assessment (RULA) from the University of Nottingham in England.

Each assessment score was weighted for statistical purposes. Because fatigue is a dynamic and cumulative process over time, a time-weighted method was used to weigh the actual duration of lifting, carrying, and static assembly.

All experiments were done in the morning to ensure consistency of the participants’ physical condition, and each experiment was only performed once a day for each participant. 

The researchers found that cerebral oxygen levels (the blue line) had a stronger correlation to physical fatigue (the orange line) than the energy expenditure model in Tecnomatix Jack software. Photo courtesy Beijing Institute of Technology

The participants—men aged 23 to 24—were divided into six groups, with three workers in each group. Workers were asked to lift weights of 8, 14 or 20 kilograms to a height of 15 or 45 centimeters from the ground. While wearing augmented reality goggles, workers had to pick up a weight, move it 3 meters, and “install” it in a virtual vehicle body. The latter process required bending over to attach the seat to the vehicle, during which time their bodies remain relatively static and only their hands perform the task.

The level of physical fatigue was represented by three indicators: cerebral oxygen saturation, physical fatigue evaluation, and metabolic energy expenditure. Our hypothesis was that fatigue would vary depending on the task. For example, lifting a 20-kilogram weight to a height of 45 centimeters would be more fatiguing than lifting a 8-kilogram weight to a height of 15 centimeters.

To measure cerebral oxygen levels, we used wireless blood oxygen sensors attached to a headband. The Augmented Reality-based Ergonomic Platform (ARE Platform) was used to create a consistent virtual assembly application. The platform puts virtual manufacturing models in a physical environment, providing participants with a semi-immersive working experience and retaining the constraints of the physical environment. A motion capture system collected data for the ergonomic indexes (RULA, OWAS, and NIOSH).

Results

Our results showed that when the lifting height was 15 centimeters and the weight was 20 kilograms, the fatigue level was the highest. On the contrary, when the height was 45 centimeters and the weight was 8 kilograms, the fatigue level was the lowest. 

The correlation between the physical fatigue evaluation method and cerebral oxygen levels was slightly higher than that between the metabolic energy expenditure estimates in Tecnomatix Jack and cerebral oxygen levels. In other words, our proposed cerebral oxygen method is more accurate than Tecnomatix Jack at predicting physical fatigue in workers.

Our results point to a new way to measure fatigue using wearable blood oxygen sensors. In recent years, there have been many cases of using wearable devices for fatigue evaluation, and most of them have also achieved good results. The use of wearable devices has the advantages of high precision, high flexibility and high adaptability. In the future it may become the primary means of accurately assessing physical fatigue.

Editor’s note: This article is a summary of a research paper co-authored by Chaoran Wang, Yaoguang Hu and Tianxin Gao of the Beijing Institute of Technology, and Wanting Mao of the Imperial College of London. To read the entire paper, click here.

ASSEMBLY ONLINE

For more information on ergonomics, read these articles:
Anti-Fatigue Mats Keep Workers on Their Feet
Exoskeletons Aid Assemblers at Truck Plant
Accommodating Older Workers