Uncertainty aware BIM and extended reality construction quality inspection using Monte Carlo conformity assessment

Abstract Building Information Modeling (BIM) can connect construction components, responsibilities, inspection criteria, and quality records, but augmented and mixed reality overlays introduce spatial uncertainty that may alter conformity decisions. This study presents a methodological, pre-experimental uncertainty framework for BIM–extended reality (XR) construction inspection. It extends a product–organization–process model with machine-readable rules, registration-error propagation, and three-state pass, review-required, and fail logic. A reproducible Monte Carlo experiment simulated 200,000 one-dimensional component deviations under an illustrative ± 20 mm tolerance, with field-measurement uncertainty fixed at 1.5 mm and registration standard deviation varied from 0 to 20 mm. Deterministic false-pass rates increased from 4.13 to 30.92%. A nominal 95% guard band kept false-pass estimates below about 0.4% under the simulated conditions but routed progressively more observations to review; near a combined uncertainty of one quarter of the tolerance, about half required independent verification. The method therefore trades false-pass risk against decision coverage. These values are conditional on the synthetic defect mixture, independent Gaussian errors, one-dimensional tolerance, and selected coverage factor and do not constitute a device or field benchmark. The framework provides a quantitative basis for deciding when calibrated AR/MR overlays may support conformity assessment and when physical measurement is required.