Data corresponding to paper: "A Feasibility Study of a Hybrid Digital Shadow System for Pavement-Load Monitoring" by Crampen et al. (submitted 2026)

This data publication contains the data related to the scientific contribution "A Feasibility Study of a Hybrid Digital Shadow System for Pavement-Load Monitoring" by Crampen et al. (submitted 2026).

Abstract of the corresponding paper: Highway management requires integrated digital approaches to support predictive maintenance and extend pavement service life. Yet traffic loads, pavement-response simulations, and infrastructure context are commonly handled in isolated systems. Digital Twins can connect perception, analysis, and simulation in real-time but require heterogeneous data and models to be temporally aligned and coherently visualized.

We present a functional hybrid Digital Shadow demonstrator for road pavement-load monitoring that establishes interoperability between matching laboratory-scale and highway field installation physical twins, integrating both domains within a shared virtual orchestrator. The system integrates pressure sensing, optical vehicle tracking, chassis sensing, weigh-in-motion, finite element simulations, and a non-intrusive surrogate model into a georeferenced environment, enabling measurements and simulation results to be inspected in spatial and temporal context.

Live data integration and real-time pavement-response prediction are demonstrated in the laboratory, while weigh-in-motion (WIM) and chassis-sensor measurements are incorporated through timestamp-consistent replay, and high-fidelity pavement finite element (FE) simulations are integrated offline. With six simultaneously represented vehicles, the platform maintains 85 frames per second (FPS). Average round-trip latency remains below 10 ms and jitter below 6 ms over a wireless local area network (WLAN) at tested send intervals of 2 to 50 ms. The surrogate predicts pavement-response fields in approximately 10 ms, reducing computation time by three orders of magnitude versus the full-order FE model, and achieves a normalized mean squared error (NMSE) below 3 % for 80 % of 1,480 evaluated pressure states. Although pavement-response validation remains model-to-model, the demonstrator provides a functional, extensible foundation for future operational road Digital Twins.