Introduction
Safety has been one of the main concerns of the construction industry in recent years. Construction fatalities, injuries, and accidents impose irreparable losses and enormous costs on this industry every year. Despite all the advances in new methods, tools, equipment, and regulations, accident rates—particularly for the four leading hazards—have not improved considerably. The four leading fatal hazards in the construction industry are falls from height, being caught between machinery or fixed structures, being struck by objects, and electrocution. Among these, falls from height were selected as the subject of this study.
Persistent safety problems have prompted safety professionals and researchers to explore more innovative ways to address the issue. One area of safety is training. Continuous yet effective training can provide new employees with essential safety knowledge and update the knowledge of current employees. However, such training sessions require budget, time commitment, and logistics. While all employees typically receive safety instructions at the outset and throughout their work, frequent review of the content is beneficial. Nevertheless, the nature of safety instructions may be tedious for some, and therefore fails to capture employees' attention, especially among younger workers. In recent years, virtual reality (VR) has facilitated safety training.
This article briefly discusses the design and development of a VR model for safety training in roofing operations. The VR application was designed and developed using an agile method. The main objective of this study is to determine the extent to which roofing professionals regard a VR application as a supplementary tool for improving their safety knowledge. The project aims to keep users engaged throughout the application while safety instructions are delivered. To present the safety content in a practical manner, a professional expert advisory group was established to provide feedback at various stages of the design and, later, the development process. This assured the design and development teams that accurate information was presented and simulated in the VR application.
After the model was completed, a quantitative approach was employed to analyze the data and examine various aspects of the VR model. To this end, a questionnaire was designed and distributed among roofing professionals to explore their opinions on the use of VR for safety training in their field of expertise. Figure 1 shows an overview of the research methodology.
Research model
The project's research methodology was designed based on the concept of progressive elaboration, as the development team continuously used industry input as the primary source for scenario development and validation. Following project initiation and an initial review, the project structure was defined. In the project development phase, the data model was created and the initial data were analyzed. Next, the VR application settings were examined and the optimal configurations for hardware, software, and the application were defined. To optimize the development process and enable mass use, the Oculus Quest 2 was selected as the VR headset for the application. Among game engines, Unity was deemed suitable for this type of project. In the next step, and through interaction with industry experts, action plans were formulated. The action plans include safety regulations and instructions across various scenarios. Ten separate sections consisting of ladder-related tasks were individually designed and developed. The main topics of these sections include:
- assessing the worksite for the best ladder placement,
- selecting the ladder to use based on height rating and building weight,
- inspecting the ladder for grease, mud, structural damage, and missing parts,
- setting up the ladder,
- securing the ladder,
- maintaining safe contact with the ladder,
- using the ladder to transport materials,
- dismounting the ladder onto the roof,
- climbing down the ladder and lowering it to the ground,
- roofing safety inspection, and
- loading the ladder onto a truck for safe transport.
In each section, it was assumed that the user was a newly hired roofing worker who had to assess safety concerns and hazards at various stages of ladder-related tasks. Figure 2 presents images of various features of the VR application that the trainee experiences during training using the VR headset.
Images of the VR application
Testing, Validation, and Results
After each section was developed, it was tested by industry experts to ensure the applicability and accuracy of the embedded components and required actions. The experts were introduced to the research project by the sponsoring organization, which is the leading professional organization for roofing contractors in the United States. During the project initiation phase, one of the points the researchers and the sponsor agreed upon was the full participation of the sponsor's representatives in the design and development processes. This led to the formation of a working group that was involved from the beginning of the research project to its end. They acted as the quality control unit in the design and development processes. This process was conducted in addition to the internal review process to ensure the validity of the VR application. Once the beta version of the application was ready, it was made available for general user feedback. Members of the industry working group invited their employees to use the application and share their opinions. A survey link was provided to users for follow-up and answering questions. The questions presented in the survey covered core content related to traditional safety training and the application of VR technology as a supplementary tool. A quantitative method was chosen for collecting data and user experience. Following the demographic section, a five-point Likert scale was used to measure levels of agreement or the magnitude of effects. Responses were collected over a 10-day period. After data collection, data cleaning was performed and the data were modeled in statistical software. In total, 46 valid samples were used in the analysis. Cronbach's alpha was used to examine the consistency and reliability of the survey. Cronbach's alpha values of 0.80 and 0.89 were obtained from the data analysis for the influencing factors and VR features sections, respectively, classifying the internal consistency of the instrument as very good.
Among the 46 collected samples, most respondents were male, over 51 years old, and had more than 25 years of work experience. Respondents were also mostly senior managers of companies with 51 to 200 employees. Following the demographic questions, participants were asked to identify the main cause of accidents in the roofing industry. While worker carelessness accounted for a high percentage, 42 percent of participants believed that a lack of knowledge or ineffective training was the cause of accidents. Participants also used a five-point Likert scale (very low: 1; very high: 5) to assess the extent to which various factors influence accidents in the roofing industry.
The next section of the survey included questions about virtual reality and its application for safety training. First, participants were asked to rate their overall familiarity with virtual reality using a five-point Likert scale. In the next question, participants expressed their level of agreement with the statement that “virtual reality is highly applicable for workforce training in roofing.” While 39% of participants agreed with this statement, 53% were neutral, and only 8% disagreed with the applicability of VR in safety training. Similarly, participants were asked whether they believed virtual reality could help the new generation of the workforce acquire the necessary safety instructions while enjoying the VR environment. The majority of participants (55% agreeing, 37% neutral, and 8% disagreeing) showed a positive perception of the effectiveness of VR for training the new generation. Participants also expressed their intention to seek opportunities to use VR applications for safety training in their companies. While 45% of participants indicated their agreement with using VR for training purposes, 44% were neutral, and only 11% of participants disagreed with this approach. Furthermore, after taking part in the designed safety program, respondents showed a positive reaction to the VR application. They were asked whether their confidence in the use and application of virtual reality had increased after participating in the program. Participants were also presented with various VR features to rate using a five-point Likert scale. While all features were rated favorably (above the midpoint), the three features “ease of use,” “comprehensive and precise instructions,” and “engaging audio/visual details” ranked among the top three desired features. Finally, participants were asked to identify the biggest barrier to the mass adoption of virtual reality technology in the roofing industry, with five common barriers offered as choices. “Consumer and business reluctance to adopt VR” was selected as the main barrier (30%).
Conclusion
Overall, the results indicated a positive attitude among professionals toward the application of VR for safety training, and it was shown that the use of VR applications as a supplementary training tool has a positive effect on the perceptions of roofing professionals. The results also showed that the influencing factors identified in this research can be taken into account in the design and development of future VR-based safety applications.
It is worth noting that while the findings of this study demonstrate the positive role of VR applications, the results are not necessarily generalizable. A larger sample size with varying levels of familiarity with virtual reality would increase the reliability of the research. In addition, classifying individuals by age, level of safety expertise, company size, positions, and traditional safety training routines would reveal further insights in future studies.