Ground-penetrating radar (GPR) has been implemented as a promising sensing technique for intelligent compaction for rapid, non-contact, and continuous estimation of asphalt pavement dielectric properties during construction. Since dielectric constant is strongly correlated with asphalt mixture density, GPR measurements can be used to infer the spatial distribution of compaction quality. However, field implementation remains challenging because signal stability is easily degraded by water films on the pavement surface and by antenna vibration under roller operating conditions. Existing studies have also focused primarily on laboratory-scale dielectric–density relationships and density prediction, with limited attention to an integrated field-oriented framework for real-time processing and visualization.

This study presents an automated GPR-based system for real-time compaction quality evaluation and visualization in asphalt pavement construction. The system includes hardware selection based on field requirements for data quality, spatial resolution, and real-time response; a modified mounting structure that improves signal stability under strong vibration; and a synchronization strategy that combines timestamps with path interpolation to align GPR and GPS-RTK data. A standardized processing workflow based on the surface reflection method was established to calibrate, detect, and correct dielectric constant measurements under the coupled effects of surface water and mechanical vibration. An automated software platform was developed to realize real-time denoising, spatial mapping of the construction area using an Alpha-Shape-constrained method, repeated compaction identification with temporal updating, and dynamic generation of compaction and temperature heatmaps.

Field experiments conducted on a highway project in China demonstrated the stability, timeliness, and engineering applicability of the proposed system. The measured average PWL ranged from 96.4 to 97.4, and the generated heatmaps effectively captured the spatial distribution characteristics of compaction quality and temperature. The proposed method provides a practical framework for real-time field compaction quality control.