Evaluation of Construction Workers' Emotional States during Virtual Reality-Based Safety Training

Abstract

Safety training is essential to encouraging construction workers to actively prevent construction accidents. Virtual reality (VR)-based safety training is becoming popular because it enhances trainees' learning without increasing physical hazards. Investigating the effectiveness of VR-based safety training will provide a basis to determine what should be maintained and complemented in safety training. A wearable electroencephalogram (EEG) device can analyze trainees' valence and arousal during the training and, in turn, measure their emotions (e.g., excitement and disappointment) which indicate their interests in learning and their level of knowledge acquisition. Thus, the emotional states can be used to evaluate the training effectiveness. Unfortunately, it still remains unclear how trainees' emotions vary throughout the training. Therefore, the authors measure construction workers' emotions during a VR-based safety training using EEG signals to evaluate its effectiveness. To achieve this objective, the VR safety training regarding fall prevention was performed with 20 construction workers who wore a wearable EEG headset to collect their EEG signals during the training. The training results are threefold. First, while learning about a safe harness and lanyard, an average construction worker felt excited except for when a scaffold system was collapsed. Second, experienced construction workers on average felt excited throughout the training. Third, while less experienced workers felt excited during the introduction of the safety gears, they felt tense when making a safety decision. These results indicate that using EEG signals during VR-based safety training will enable trainees' emotional state measurement and, in turn, a better design of safety training.