Jongwon Seo

With the rapid advancement in information technology, the digital twin and smart assembly process simulation have become an integral part of the design and manufacturing of high-precision products. However, conventional Tunnel Boring Machine (TBM) cutterhead design and on-site assembly planning remain largely experience-driven and fragmented, with limited interoperability between geological characterization, structural verification, and constructability validation. This study proposes a digital twin-driven framework for TBM cutterhead design optimization and assembly process simulation that integrates geology-aware design inputs, BIM-based information modelling, FEM-based structural assessment, and immersive virtual environments within a unified virtual–physical workflow. To ensure consistent data exchange across platforms, an IFC4.3-compliant ontology is established using a non-intrusive property-set (Pset) extension strategy to represent cutterhead components, geological parameters, FEM load cases/results, and assembly tasks. Tunnel-scale stress analysis and cutter–rock interaction modelling are used to define project-representative cutter loading envelopes, which are mapped to a high-fidelity cutterhead FEM model for iterative structural refinement. The optimized configuration is then transferred to a game-engine/VR environment to support full-scale design inspection and assembly rehearsal, followed by manufacturing and field deployment with bidirectional feedback. To validate the proposed framework, an implementation case study of a deep hard-rock tunnelling project is presented where five design iterations were tracked across BIM–FEM–VR and nine constructability issues detected and resolved prior to assembly. The results indicate that the proposed digital twin approach strengthens traceability from geology to loading to structural response, reduces localized stress concentration at critical interfaces, and improves assembly readiness for complex tunnelling projects.

Earthwork operations constitute a substantial share of infrastructure project costs and are critical to overall project efficiency. However, the construction industry still relies on conventional approaches and there is a lack of integrated fleet management systems for collaboratively working equipment. While telematics is widely used in other industries, its applications to monitor the complex interactions between excavators, dump trucks, and dozers in real time remain limited. This study proposes an intelligent fleet monitoring system that utilizes only satellite navigation data (GNSS) to analyze the real-time productivity of multiple earthwork machines without relying on additional sensors, such as IMU or accelerometers, thereby eliminating the need for separate measurement procedures. A lightweight site configuration step is required to define the work area/loading/dumping geofences on an existing site map. This research provides novel developed algorithms that facilitate a real-time productivity assessment for several earthwork equipment and provide planning-level recommendations for equipment deployment combinations. Dedicated motion classification algorithms were developed for excavators, dump trucks, and dozers to distinguish activity states, to compute working and idle times, and to quantify operational efficiency. The system integrates a web-based e-Fleet Management platform and a mobile e-Map application for visualization and equipment optimization. Field validation was conducted on two active earthwork projects to evaluate accuracy and feasibility. The results demonstrate that the developed algorithms achieved classification and productivity estimation errors within 2.5%, while enabling optimized equipment combinations and improved cycle time efficiency. The proposed system offers a practical, sensor-independent approach for enhancing productivity monitoring, real-time decision-making, and cost efficiency in large-scale earthwork operations.

Recently, deep learning algorithms have been increasingly applied in construction for activity recognition, particularly for excavators, to automate processes and enhance safety and productivity through continuous monitoring of earthmoving activities. These deep learning algorithms analyze construction videos to classify excavator activities for earthmoving purposes. However, previous studies have solely focused on single-source external videos, which limits the activity recognition capabilities of the deep learning algorithm. This paper introduces a novel multi-modal deep learning-based methodology for recognizing excavator activities, utilizing multi-stream input data. It processes point clouds and RGB images using the two-stream long short-term memory convolutional neural network (CNN-LSTM) method to extract spatiotemporal features, enabling the recognition of excavator activities. A comprehensive dataset comprising 495,000 video frames of synchronized RGB and point cloud data was collected across multiple construction sites under varying conditions. The dataset encompasses five key excavator activities: Approach, Digging, Dumping, Idle, and Leveling. To assess the effectiveness of the proposed method, the performance of the two-stream CNN-LSTM architecture is compared with that of single-stream CNN-LSTM models on the same RGB and point cloud datasets, separately. The results demonstrate that the proposed multi-stream approach achieved an accuracy of 94.67%, outperforming existing state-of-the-art single-stream models, which achieved 90.67% accuracy for the RGB-based model and 92.00% for the point cloud-based model. These findings underscore the potential of the proposed activity recognition method, making it highly effective for automatic real-time monitoring of excavator activities, thereby laying the groundwork for future integration into digital twin systems for proactive maintenance and intelligent equipment management.

Earthwork construction presents significant challenges due to its unique characteristics, including irregular topography, inhomogeneous geotechnical properties, dynamic operations involving heavy equipment, and continuous terrain updates over time. Existing methods often fail to accurately capture these complexities, support semantic attributes, simulate realistic equipment–environment interactions, and update the model dynamically during construction. Moreover, most current digital solutions lack an integrated framework capable of linking geotechnical semantics with construction progress in a continuously evolving terrain. This study introduces a novel, voxel-based digital twin framework tailored for earthwork construction. Unlike previous studies that relied on surface, mesh, or layer-based representations, our approach leverages semantically enriched voxelization to encode spatial, material, and behavioral attributes at a high resolution. The proposed framework connects the physical and digital representations of the earthwork environment and is structured into five modules. The data acquisition module gathers terrain, geotechnical, design, and construction data. Virtual models are created for the earthwork in as-planned and as-built models. The digital twin core module utilizes voxels to create a realistic earthwork environment that integrates the as-planned and as-built models, facilitating model–equipment interaction and updating models for progress monitoring. The visualization and simulation module enables model–equipment interaction based on evolving as-built conditions. Finally, the monitoring and analysis module provides volumetric progress insights, semantic material information, and excavation tracking. The key innovation of this framework lies in multi-resolution voxel modeling, semantic mapping of geotechnical properties, and supporting dynamic updates during ongoing construction, enabling model–equipment interaction and material-specific construction progress monitoring. The framework is validated through real-world case studies, demonstrating its effectiveness in providing realistic representations, model–equipment interactions, and supporting progress information and operational insights.

The construction industry faces challenges of labor shortages and safety concerns. Machine control (MC) technology offers a solution, particularly for graders in earthmoving operations. This study introduces a path-planning algorithm using initial site data, 3D target models, and equipment specifications to create optimal work paths for graders. The algorithm minimizes data exchange and adjusts for varying road widths and curves, enhancing grading efficiency and accuracy. A case study on a road construction site in South Korea validated the algorithm’s practical application. The proposed system aims to improve construction efficiency, safety, and cost-effectiveness, contributing to the advancement of construction automation technology.

Excavator–truck operations, characterized by their repetitive excavation and loading tasks, present a prime candidate for automation. While numerous studies have aimed to automate the earthworks, practical implementations remain scarce. This research introduces a task component design focused on excavator–truck operation planning to improve the functionality of an earthwork automation system. To address this, fundamental task primitives necessary for executing excavation tasks were engineered, and a web-based system was developed to automate the generation of work plans for both point and trench excavation through algorithmic processes. Additionally, a JSON-based protocol was introduced to facilitate efficient integration with other subsystems. Field tests were conducted to validate the effectiveness of the newly developed algorithm and protocol within the broader context of earthwork automation systems. The results demonstrated the successful implementation of the task components, confirming their operational viability and seamless integration into the existing earthwork automation framework.

Currently, earthworks expenses, which account for a large portion of civil engineering work, are managed by hand or offline, such as invoices. However, such manual management can lead to human errors and poor management and manipulation of records, resulting in difficult situations that cause disputes between the client, construction company, partner company, and equipment technicians. In addition, a system that can transparently manage sand and soil is needed as illegal cases occur frequently, such as randomly bringing in rocks and soil at construction sites or unauthorized discharge and landfill of defective soil. In this paper, we want to use a Web-GIS-based smart invoice system to secure transparency and efficiency in bringing in external soil, identify the operation path and time of construction equipment with a GPS-based transport management system, and analyze the details by comparing contract details and big data. In addition, we intend to contribute to improving safety management by raising awareness of safety accidents through inspection and regulation of speeding vehicles using smart invoices. By calculating the optimal number of Sato equipment, we intend to increase system productivity by minimizing idle time of equipment and to summarize improvements by deriving smart invoice problems.

The rapid development in underground infrastructure is encouraging faster and more modern ways, such as TBM tunneling, to meet the needs of the world. However, tunneling activities generate complex and heterogeneous data, which makes it difficult to visualize the performance of a project. Advancements in information technology, such as digital twins and machine learning, provide platforms for digital demonstration, visualization, and system performance monitoring of such data. Therefore, this study proposes a digital twin-driven framework for TBM performance prediction through machine learning, visualization, and monitoring. This novel approach integrates machine learning and real-time performance data to predict, visualize, and monitor the status of the tunnel construction progress. A digital twin virtual model of TBM was constructed based on TBM design parameters, the input parameter, boring energy, RPM, torque, thrust force, speed, gripper pressure, total revolution, and Q-value provided to SVR and ANN models to predict the TBM AR and PR, and TBM daily progress was visualized continuously. The predictive performance indices R² (0.97) and RMSE (0.011) were estimated for AR prediction, showing the accuracy of the proposed model. To demonstrate the proposed framework, this study shows the its effectiveness. By implementing this framework, stakeholders can minimize the risk associated with the cost and schedule of a tunneling project by simultaneously visualizing and monitoring the performance of TBMs through digital twin and machine learning algorithms.