This study aims to evaluate the effectiveness of an augmented reality-based just-in-time (JIT) information presentation system (AR-JIT system) in supporting construction task performance, rework outcomes and safety awareness compared to traditional paper-based methods.
A total of 40 participants were randomly assigned to either an AR-JIT system or a paper-based condition in a controlled experiment. They completed a standardized rebar installation task measuring spatial reasoning and safety. Key outcomes included task accuracy, completion time, rework frequency, cognitive workload and hazard identification. Statistical and factorial analyses compared group performance and assessed demographic effects.
Participants using the AR-JIT system achieved significantly higher task accuracy than those using paper-based methods (89.0% vs 73.5%, p < 0.001). The AR-JIT system demonstrated a substantial reduction in mean rework frequency relative to the paper-based method (65.6% lower), with a moderate effect size, although this difference did not reach statistical significance in the present sample. No statistically significant differences were observed between groups for cognitive workload (p = 0.093), task completion time (p = 0.465) or safety hazard identification performance (p = 0.271). Demographic factors, including age, gender and prior construction experience, did not show statistically significant interactions with information delivery method for the primary performance outcomes.
The laboratory setting, single-task focus and predominantly student sample limit generalizability to real-world construction environments. Future studies should examine AR-JIT systems in field settings with larger samples and longitudinal designs.
This study offers experimental evidence that a marker-based AR JIT information system can enhance task accuracy in construction without adding time or cognitive demands.
