While evaluating the physical workload on workers, the force exerted needs to be estimated in a convenient way because biomechanical analysis for physical workload evaluation requires data on both posture and exerted force. The aim of this study was to investigate the effectiveness of exerted force estimation using a wearable device that can be easily utilized even by ordinary workers. We measured eight electromyograms (EMGs) on the forearm and the forearm posture with an armband type wearable sensor during manual material handling with varying holding weights and heights. The measurement results showed that the EMGs monotonically increased with an increase in weight. In addition, the EMGs varied with the holding height even when the same weight was held. We constructed an estimation function of the weight using multiple regression analysis. Two sets of explanatory variables were used to investigate the effectiveness of the forearm posture data: the eight EMGs (i.e., SET 1) and the eight EMGs and the two forearm angles (i.e., SET 2). Multiple regression analysis showed that the accuracy of SET 2 was better than that of SET 1. In addition, the average absolute error of the estimation function with SET 2 was 1.49 kg; thus, we concluded that the accuracy of this estimation function has sufficient accuracy for the evaluation of physical workload.

中文翻译：

---

在手动物料搬运过程中使用可穿戴传感器估算力

在评估工人的身体工作量时，需要以方便的方式估算所施加的力，因为用于物理工作量评估的生物力学分析需要有关姿势和所施加力的数据。这项研究的目的是研究使用即使普通工人也可以轻松使用的可穿戴设备进行力估算的有效性。我们在进行人工物料搬运时，使用腕带式可穿戴传感器在前臂和前臂姿势上测量了八张肌电图（EMG），握持重量和高度各不相同。测量结果表明，肌电图随重量的增加而单调增加。另外，即使保持相同的重量，EMG也随着保持高度而变化。我们使用多元回归分析构建了权重的估计函数。两组解释变量用于研究前臂姿势数据的有效性：八个EMG（即SET 1），八个EMG和两个前臂角度（即SET 2）。多元回归分析表明，SET 2的精度优于SET1。此外，SET 2估计函数的平均绝对误差为1.49 kg； SET 2的估计函数的平均绝对误差为1.49 kg。因此，我们得出结论，该估计函数的准确性对于评估物理工作量具有足够的准确性。SET 2的估计函数的平均绝对误差为1.49 kg；因此，我们得出结论，该估计函数的准确性对于评估物理工作量具有足够的准确性。SET 2的估计函数的平均绝对误差为1.49 kg；因此，我们得出结论，该估计函数的准确性对于评估物理工作量具有足够的准确性。