Risk awareness assessment of construction workers wearing AR head-mounted displays using EEG signals

Augmented reality (AR) head-mounted displays have gained attention in the construction industry for enhancing information visualization and operational efficiency. However, their impact on workers' risk awareness remains underexplored.

This study proposes a novel EEG-based framework to evaluate the risk awareness of construction workers performing tasks in an AR-enhanced environment. Forty participants equipped with AR headsets and EEG devices completed a series of tasks-standing, walking, and measuring-on a simulated elevated steel frame. EEG signals were processed using wavelet transforms to extract real-time power in theta, alpha, beta, and gamma bands, from which engagement (ENI) and vigilance indices were derived.

The result shows that cognitive engagement increases while vigilance decreases during high workload tasks such as measuring, indicating reduced risk awareness. Statistical analyses confirm that task type significantly affects both ENI and vigilance, while repeated exposure improves overall vigilance but has limited effect on high-load tasks.

These findings provide a neurocognitive perspective on how AR affects worker safety and highlight the importance of task-specific design and training in AR-integrated construction work.