Learning Where to Discover Safety Hazards: Modelling Eye-movement to Train Construction Students' Hazards-identification Ability

Project: Research

View graph of relations


The construction industry is hazardous due to the congested working environment andcomplexity (Figure 1). Hazard identification is the most frequently employed approach topreventing and reducing accidents on construction sites. However, hazard-identification is acomplex task that requires knowledge of both regulations and experience, and the hazardidentificationexperience is a kind of tacit knowledge that is difficult to be described explicitlyin the text.Researchers from other fields explored applying eye-movement modeling examples (EMME) topresent experts' experiences in training materials. EMME is a video (a screenshot presented infigure 2) showing how experts perform a task, including experts' gaze points and voice whichexplicitly express experts' sensory process and cognitive process, respectively. It has been appliedfor education in several disciplines except the construction training, e.g., medical training andcomputer programming education. Previous findings show it can lead to better learning outcomesas well as changes to learners' eye-movement. Besides, according to the Cognitive Theory ofMultimedia Learning, students apply both their visual and audio channels when learning fromthe EMME, so the EMME is possible to improve learning outcomes. As experts are supposed toallocate their attention in task-related areas, learners receive less redundant information duringlearning. Even though the EMME is beneficial in education, the construction industry has notexplored EMME-related applications yet to improve the issues.Hazards-identification can be regarded as a visual search task requiring experiences. It is riskyand costly to bring students to real job sites to train them for hazard identification; therefore, thisproposed project intends to explore the EMME application to enable students to experience theactual job site situations for improving the construction training. This project aims to explore theeffectiveness of using EMME in training hazards-identification and develop a series of eyemovementmodeling examples that effectively help construction students learn hazardsidentification.Preliminary EMME will be shot in various construction locations first, andconstruction students will be invited to attend a pre-test to examine their original hazardsidentificationability. Then those students will be required to study the preliminary EMME. Next,they will join a post-test to assess their hazards-identification ability after learning from theexamples. Finally, the examples will be modified based on participants' feedback and changes.Meantime, several experts from the construction industry and academia from universities will beinvited to optimize the modified examples further. The final examples will be cut into multipleshort videos according to the construction stages and locations.


Project number6000770
Grant typeTDG(CityU)
Effective start/end date1/06/2130/09/23