Abstract
Abstract Introduction The quality assessment of tourism highway design plays a pivotal role in achiving a harmonious balance among transportation functionality, drivers’ perceptual experiences, and ecological coordination. Traditional evaluation methods, which predominantly rely on post-construction expert evaluations or static visual surveys, exhibit limitations in terms of spatial coverage, repeatability, and the capacity to capture drivers’ continuous, and dynamic responses to the roadside landscape. Objectives To address the above limitations, this study aims to establish a human-factor-driven evaluation framework of Building Information Modeling (BIM)-based digital twin models and immersive virtual driving simulations for tourism highways. Methods A perception-driven digital twin approach was used to precisely replicate road infrastructure, terrain, and auxiliary facilities in a virtual environment by fusing multi-source data from BIM, Generalized Information System (GIS), oblique photogrammetry, and a Internet of Things (IoT) layer. The virtual driving system incorporated multiple functional modules, e.g., immersive driving simulations, dynamic weather simulations, and synchronized physiological data acquisitions. And then, 50 participants were recruited to drive in the virtual S103 tourism highway under three weather conditions with the collection of heart rate variability (HRV), stress, skin conductance level (SCL), and speed variability (VSD). Results Experimental results demonstrated spatially clustered physiological responses that reduced HRV and elevated stress appeared consistently at specific road segments. The skin conductance captured moment-to-moment emotional fluctuations, and VSD revealed individual differences in driving stability. The multi-indicator data successfully identified dangerous segments with high driver workload and provided a cross-validated human-factor profile along the entire route. Conclusion The multi-dimensional human-factor evaluation framework of BIM digital twins and immersive virtual driving for tourism highways developed in this study was proved to be effective and helpful in pre-construction design diagnosis, quantitative comparison of design alternatives, and evidence-based optimization of landscape rhythm and facility layout. The approach enhances the objectivity, repeatability, and perception-centered nature of tourism highway evaluation and holds the potential of a routine design-aiding tool in future highway engineering applications.