The use of digital twins and geographic information systems to optimize logistics in the construction of high-speed highways

Abstract

The study seeks to substantiate and quantify the effect of integrating digital twins and geographic information systems to optimise logistics in the construction of high-speed rail lines against the backdrop of the sector’s chronic productivity crisis and the high capital intensity of megaprojects, which makes the reduction of timelines and budget overruns critically important; the focus is the transformation of the logistics subsystem, as supply disruptions trigger cascades of delays and idle time on sites, amplifying variability and managerial uncertainty. The aim is to empirically compare traditional logistics management (TLM) with an integrated DT GIS Optimised model (DGO) and to measure its impact on logistics, production and economic KPIs under realistic constraints. The methodological basis combines a synthesis of peer-reviewed literature and industry reports with conceptual simulation modelling of a scenario involving the construction of a 150 km HSR segment with the delivery of 50,000 precast segments, 2 million m³ of concrete and 300,000 t of rebar to 25 parallel sites; the DGO architecture integrates BIM, IoT (GPS, RFID) and GIS via a digital-twin middleware core with the exchange of actionable intelligence in lightweight formats (JSON) for dynamic routing and proactive resource reallocation. The results show an increase in on-time-in-full (OTIF) from 82,47% to 98,92%, an 80,24% reduction in delivery time variance, a 31,60% reduction in on-site vehicle turnaround, and a 14,97% decrease in cost per tonne kilometre; on site, equipment downtime due to material shortages decreases from 18,26% to 2,58%, labour productivity rises by 28,41%, and daily schedule adherence reaches 95,71%, enabling a 63,79% reduction in the time buffer. The economic block records initial investment of 2,75 million c.u., net annual savings of 2,60 million c.u., and a payback period of 12,69 months, with the principal effect linked to a radical reduction in variability and a shift to predictive real-time management. Limitations concern cybersecurity, workforce retraining and organisational resistance; prospects include ML/AI integration for autonomous dispatching and the expansion of the digital twin to the full asset life cycle.