Models on eye movements, attention may reduce accidents

Human error is responsible for up to 80% of workplace accidents, but predictive models evaluating workers’ eye movements and attention may help reduce incidents, according to a study released Tuesday by researchers at George Mason University in Fairfax, Virginia.

By measuring eye movements and cognitive processes for at-risk workers in the construction industry build, the researchers built an error-detection framework that computes the likelihood of human error in occupational settings. The researchers linked eye movement and workers’ attention with research focused on working-memory load and decision making to understand how and why workers in a dynamic environment fail to detect, comprehend and/or respond to physical risks.  

The researchers found that:

• Eye movements can be used as precursors of workers’ safety errors.
• Working-memory load and personality traits play a significant role in risk-taking behavior.
• Data mining classifiers can be used to compute the likelihood of different types of human errors.  

The researchers said the use of predictive models like the one used for the study may lead to “significant accident reduction” in the workplace and enhance workers’ risk-analysis skills.