Abstract
Continuouspavementassessmentisrequiredforsustainablehighwaymanagement. Manual surveys are limited by subjectivity, delay, safety risks, and cost. Automated approaches can support objective and timely pavement assessment. This article presents an integrated method in which edge vision and global navigation satellite system (GNSS) telemetry are combined. A web geographic in- formation system (WebGIS) is also included. Pavement condition index (PCI) prediction is linked with life-cycle cost analysis (LCCA). Pavement defects are detected through a hybrid deep learning approach. Physical defect areas are estimated through calibrated image transformation. Geospatial records are transmitted to the WebGIS platform. Pavement condition is predicted through a cali- brated machine learning model. Independent test data are reserved for model evaluation. Regulatory thresholds are defined according to national pavement standards. Reactive and proactive maintenance strategies are compared over a long planning horizon. The results indicate reliable defect detection and PCI prediction. The proposed method provides rapid field assessment and geospatial report- ing. Proactive maintenance provides lower lifecycle costs than reactive maintenance. Stable economic benefits are observed across the tested sensitivity cases.