The efficacy of virtual-reality based safety training in construction: Perceptions versus observation

Abstract

The construction industry is widely regarded as a high-risk environment that is also known for its high fatalities and injury rates. The lack of appropriate health and safety training is a leading cause of accidents on construction sites. In recent years, several industries have experimented and gradually employed Virtual Reality (VR) technology to enhance safety training. Unlike traditional health and safety training that relies on a one-way instructional method of delivering information, VR provides experienced-based learning by placing the user inside a drill. This paper reports on a preliminary study that sought to assess the size of the efficacy of VR-based training within the construction industry context. By utilizing a mixed-method case-control research methodology, the study compared achievements made through the traditional methods of health and safety training such as the use of PowerPoint presentations against VR-based training. The achievements were evaluated based on the immediate perception of the trainees compared to the actual observations afterwards. The study focused on two particular aspects of the training outcomes: (1) impact on the knowledge retention by the trainees to identifying risks on construction sites, and (2) influence on their future risk-taking behavior. Eight residential construction sites in Auckland, New Zealand were selected as the indicative clusters of construction activities in the region. Correspondingly, a simple random method was applied to select one of the sites and sample ten trainees from the nominated site. Accordingly, a control group of five professionals received approximately 25 minutes of traditional health and safety training. An intervention group of another five professionals received the same length of health and safety training while using VR technology. After each training session, the groups were interviewed through the use of face to face semi-structured strategy twice. In both rounds of the interviews, the respondents were quantitatively assessed by rating their responses on a 5-point Likert scale. The first round was conducted immediately after the training session. It graded the perception of the respondents about their level of engagement and satisfaction, besides the impact of the session on their future safety behavior. In order to examine the genuine achievements of the training sessions, the second round of interviews was instituted the day after the training session. It assessed the participants' recall of the training contents and their risk-taking behavior in a hypothetical scenario. Despite the strong belief from trainees about the effectiveness of the VR technology in improving their health and safety learnings, no significant difference was observed in their risk-taking behavior and risk identification ability. Furthermore, some cases of nausea and dizziness were observed when using the VR headsets. The results can be generalized only if the future studies with bigger sample size show a similar outcome.