AI-Driven VR Training Could Cut Struck-By Incidents on Road Crews
Struck-by incidents on road construction sites killed about 1,800 workers and caused more than 167,000 nonfatal injuries between 2011 and 2021, according to National Institute for Occupational Safety and Health data cited by Construction Dive. In a Q&A with the publication, Namgyun Kim, an assistant professor of construction science at Texas A&M University, said VR- and AI-based safety training “should not be viewed as an experimental technology anymore.” Kim, working with colleagues at Texas A&M and Louisiana State University, is studying how generative AI can build project-specific training scenarios rather than relying on fixed, generic VR simulations. The research also targets “habituation,” the phenomenon where workers on busy highway jobsites gradually stop noticing recurring hazards like warning signs and alarms after prolonged exposure. Kim said AI and wearable tech can now identify these patterns and trigger personalized training before they lead to an accident.
What It Means for Subcontractors
- Highway and roadside construction crews should evaluate VR training platforms now rather than waiting for a mandate. Kim says commercial VR systems are already widely available, and AI tools can cut the time and cost of building scenarios specific to a crew’s actual site conditions and schedule.
- Firms struggling to justify safety tech spending should note Kim’s ROI argument: since safety incidents are rare and hard to link to a single training program, the business case rests on AI filling gaps traditional toolbox talks and classroom training can’t, such as detecting habituation before it causes a struck-by.
- Safety managers running toolbox talks should watch for emerging AI tools that analyze those conversations to flag communication gaps, a capability Kim says is being developed to strengthen safety culture and reduce reliance on one-size-fits-all training.
