Abstract Industry 5. 0 promotes human-robot collaboration focused on human well-being, requiring work designs that consider human factors. However, the robot fulfillment system (RMFS) introduced in order picking in logistics warehouses often force workers to crouch and perform complex processes, deviating from human-centered design. Previous studies have investigated the physical workload of workers in RMFS, and have not examined the effects of mental workload and personal characteristics on performance. To overcome this problem, this study conducts an experiment of order picking in RMFS to measure work time per order, individual characteristics of workers, and workload by heart rate data and questionnaires. As a performance indicator, we introduced a shape parameter and a scaling parameter, that approximates the work time by a gamma distribution, and examined the causal relationship using a Bayesian network. The main results showed that the work time per order increased for tasks with many tasks outside the neutral reach zone (Pattern 1) and tasks with no waiting time for shelf transfers (Pattern 2), compared to tasks with many tasks within the neutral reach zone (Pattern 0), which is the range where the workload on the worker is small. In particular, the datasets for Patterns 1 and 0 showed a significant correlation between frustration and ~ (r=0. 64). Although no causal relationship was found in a Bayesian network, the datasets for Patterns 2 and 0 showed a significant correlation between Extroversion and ~ (r=-0. 48), indicating that Extroversion may contribute to maintaining performance when working with AGVs.
Ono et al. (Sat,) studied this question.