Immersive Bridge Inspection: Quantifying the Benefits of XR for Anomaly Detection

We present an immersive bridge inspection tool as part of a digital twin for bridges that merges virtual reality (VR) with AI-assisted anomaly detection to evaluate how extended reality (XR) tools can modernize bridge inspections. Two tasks were studied: (i) relocating previously reported anomalies and (ii) detecting new surface anomalies using XR technologies. XR cues and simulated AI prompts were compared with manual inspection for five key performance indicators. Results show that XR + AI cut average detection time by 51%, boosted accuracy by 38%, lowered subjectivity by 11%, and reduced cognitive workload by 23 NASA-TLX points. A post-study survey yielded favorable usability and usefulness ratings, although some volunteers experienced VR-induced discomfort. These findings demonstrate the promise of XR and AI to accelerate, objectify, and lighten the mental load of structural assessments, while highlighting usability challenges that must be addressed for field deployment.