Probabilistic Evaluation of Measurement Uncertainty and Decision Risk in UAV-Based Dimensional Inspection

Unmanned aerial vehicles (UAVs) are increasingly used for remote dimensional inspection in transportation monitoring and infrastructure control. In such applications, measurement results are often interpreted relative to regulatory thresholds, making the reliability of inspection decisions strongly dependent on measurement uncertainty. This study presents a probabilistic framework for evaluating measurement uncertainty and decision risk in UAV-based dimensional inspection tasks. A measurement model describing uncertainty scaling with observation geometry is formulated, and the probability of exceedance relative to a regulatory limit is derived. The framework integrates probabilistic measurement modeling with a risk-based decision formulation that accounts for false-positive and false-negative inspection outcomes. The resulting integral inspection risk is analyzed for representative sensing modalities commonly used in UAV platforms, including vision-based systems, LiDAR, and radar sensors. The results demonstrate that uncertainty scaling with flight altitude significantly influences exceedance probability and decision reliability. Sensors with lower intrinsic dispersion maintain sharper threshold transitions and therefore provide more stable regulatory decisions. Sensitivity analysis further confirms that moderate variations in measurement uncertainty can substantially affect inspection risk. The proposed framework provides a quantitative tool for evaluating sensing technologies in UAV-based inspection missions and supports the design of reliable drone-assisted dimensional compliance monitoring systems.