This study integrates UAV imagery and 3D geospatial modelling into the design Value Engineering (VE) process to address long-standing limitations in information acquisition during the conventional pre-study phase. The research aims to determine whethe...
This study integrates UAV imagery and 3D geospatial modelling into the design Value Engineering (VE) process to address long-standing limitations in information acquisition during the conventional pre-study phase. The research aims to determine whether 3D geospatial data can support the generation of design alternatives and value propositions that are difficult to obtain solely through design documents and field surveys. Furthermore, it evaluates the potential of data-driven value analysis based on model accuracy verification, procedural quantification, and measurable improvements in value outcomes. Focusing on an urban road infrastructure project, the study conducts an objective assessment of how 3D geospatial modelling enhances the reliability and effectiveness of information acquisition within the VE workflow.
The developed 3D geospatial model provides an integrated representation of terrain morphology, road networks, and the spatial distribution of structures and residential areas. This model functioned as a visual analytical platform that enabled intuitive examination of geometric configurations, roadway and intersection alignments, and potential risk factors during the VE workshop. The photogrammetric outputs achieved a ratio of root-mean-square error to ground sample distance of 1.92, a mean reprojection error of 0.207px, focal length and principal point variations below 1%, and distortion coefficients converging within ±0.01, thereby meeting the relevant USGS accuracy thresholds. These results confirm that the model satisfies the relative accuracy requirements for application in 3D geospatial modelling–based design VE.
A total of 37 VE ideas were evaluated, comprising 31 derived from conventional design reviews and field observations, and 6 generated through 3D geospatial modelling. The overall cost variation was –2.151%, equivalent to a reduction of approximately KRW 4.47 billion. Among these, the six modelling-based alternatives accounted for –1.026% (around KRW 2.13 billion), representing 47.698% of the total savings. The average value improvement increased from 21.088% under the conventional approach to 28.983% when incorporating 3D geospatial information, demonstrating that the technology contributes not only to cost reduction but also to broader improvements in functionality, economic efficiency, and constructability. Representative cases—including tunnel length reduction, intersection realignment, changes in retaining wall methods, and the optimisation of TSP and exploratory horizontal boring—produced simultaneous benefits in traffic safety, construction period, and environmental performance.
Overall, the findings provide empirical evidence that 3D geospatial modelling can restructure the design VE workflow—spanning information acquisition, idea generation, alternative assessment, and value analysis—into a quantifiable, data-driven process. The study also confirms that integrating 3D geospatial data improves the reliability of economic assessments and enhances the consistency of VE decision-making. Nevertheless, the use of a single urban road project imposes limitations on the generalisability of the results. Future research should investigate quantitative relationships between 3D geospatial accuracy indicators (e.g., Root Mean Square Error, Mean reprojection error, Focal length/principal point variation, Distortion coefficients) and VE performance metrics (e.g., Coat Variation Rate to Total Project, CVRt; Value Improvement Ratio, VIR). Despite these limitations, the study lays an important foundation for data-centric design VE and offers a pathway towards an integrated smart VE framework linking 3D geospatial modelling with AI, BIM, and life-cycle costing.