Reduction of Variability in Visual Inspection Results Using Virtual Reality

Condition assessment, typically performed through visual inspection, is essential but often challenged by subjectivity and variability among inspectors. This paper presents a comparative evaluation of three inspection approaches applied to a reinforced concrete bridge in Ecuador: traditional on-site inspection, UAV-based image assessment, and immersive VR-based inspection. The analysis was carried out following the criteria established in the Guide to Determine the Condition of Bridges in Costa Rica through Visual Inspection, in which the condition rating (CP) of a bridge is classified on a scale from 1 to 6, with 1 indicating a satisfactory condition and 6 indicating unacceptable risk or imminent failure. The traditional inspection, performed by a single inspector, resulted in a CP of 5 within 50 minutes. UAV-based evaluation with five inspectors produced variable results, ranging from 4 to 6, with a standard deviation of 0.63 and an average duration of 65 minutes. In contrast, VR inspection produced consistent overall ratings of 5 with zero variability, though some variation remained in the individual element ratings, and it required only 15 minutes to complete. These results show that VR improves consistency and efficiency, supporting its use as a complementary tool in bridge assessment and management.