An EEG-Based Mental Workload Evaluation for AR Head-Mounted Display Use in Construction Assembly Tasks

Abstract

Augmented reality (AR) has become one of the most pervasive technologies for immersive applications with potential to improve worker performance in construction. However, the cognitive impact of using AR is still being studied and a comprehensive assessment of changes on user’s cognitive status is needed. This study proposed an electroencephalogram (EEG)-based mental workload evaluation method for a wood frame assembly, comparing the differences between two information display types (conformal model and paper layout). Additionally, a task performance analysis was conducted for completion rate and accuracy. The experiment recruited thirty construction students with similar backgrounds who were novice AR users. The findings indicated that AR improved the assembling efficiency with significantly higher completion (p<0.001) in comparison to the paper display. Furthermore, both mean and maximum values during the task process were analyzed to compare the difference between the displays, where AR display generated significantly less mental work in mean (p=0.007) and maximal (p=0.030) values. The comparison of the changing trends of mental workload increment in assembly revealed that AR display maintained lower mental workload and smaller fluctuation (standard deviation) during the entire assembly. The outcomes examined the usability of AR head-mounted display (HMD) in the construction assembly task with a moment-to-moment physiological evaluation of user’s mental workload. This work shows that EEG-based mental workload evaluation can be used as an effective method for revealing changes in workers’ mental status in real time during assembly, which has implications for future research in areas such as hazard recognition and situational awareness evaluation.