Search

Fatigue is a major cause of low back pain for workers in various fields, including industry and agriculture. It has a negative impact on workers’ safety, decreases their productivity, and causes a reduction in their occupational career. An exoskeleton is expected to be a solution for reducing workers’ fatigue. However, assessing the safety and effectiveness of exoskeletons, except for the direct measurement of electromyography (EMG) in the human body, is challenging in real-case scenarios. Recently, simulations have been widely used to estimate biomechanical variables. Thus, we aimed to develop a method that combines an exoskeleton model and human body simulation to evaluate the effects of exoskeletons on lumbar fatigue. The strength and tendency estimated using this method are similar to those obtained from EMG devices in symmetrical repetitive lifting tasks. In addition, this method can be used to predict and simulate fatigue after a recorded motion. Our findings will help guide manufacturers in designing their products.

Exoskeletons can protect users’ lumbar spine and reduce the risk of low back injury during manual lifting tasks. Although many exoskeletons have been developed, their adoptability is limited by their task- and movement-specific effects on reducing burden. Many studies have evaluated the safety and effectiveness of an exoskeleton using the peak/mean values of biomechanical variables, whereas the performance of the exoskeleton at other time points of the movement has not been investigated in detail. A functional analysis, which presents discrete time-series data as continuous functions, makes it possible to highlight the features of the movement waveform and determine the difference in each variable at each time point. This study investigated an assessment method for exoskeletons based on functional ANOVA, which made it possible to quantify the differences in the biomechanical variables throughout the movement when using an exoskeleton. Additionally, we developed a method based on the interpolation technique to estimate the assistive torque of an exoskeleton. Ten men lifted a 10-kg box under symmetric and asymmetric conditions five times each. Lumbar load was significantly reduced during all phases (flexion, lifting, and laying) under both conditions. Additionally, reductions in kinematic variables were observed, indicating the exoskeleton’s impact on motion restrictions. Moreover, the overlap F-ratio curves of the lumbar load and kinematic variables imply that exoskeletons reduce the lumbar load by restricting the kinematic variables. The results suggested that at smaller trunk angles (