Recently, research has been performed on Geographic Information Systems (GIS) to effectively recycle abundant Building Information Modeling (BIM). However, GIS and BIM use different schemas to describe forms. In particular, GIS defines the Level of Detail (LOD) in schema to quickly visualize numerous city-level objects. Because of schema differences, a mapping process is essential to reuse BIM in GIS. The problem is that mapping LOD models requires significant calculation time. This method reduces BIM-GIS mapping operation productivity, making it difficult to reuse BIM. In this research, we enhance mapping work productivity by developing a method that automates LOD geometry mapping using Screen-Buffer scanning-based Multiprocessing (SB-MP) including semantic mapping rule. The proposed method can dramatically improve LOD geometry modeling productivity. Finally, the developed results are evaluated quantitatively.

This study introduces the BIM (building information modeling)-GIS (geographic information system) conceptual mapping (B2GM) standard ISO N19166 and proposes a mapping mechanism. In addition, the major issues concerning BIM-GIS integration, and the considerations that it requires, are discussed. The B2GM is currently being standardized by the spatial-information international standardization organization TC211. Previous studies on BIM-GIS integration seem to focus on the integration of different types of model schemas and on the implementation of service interfaces. B2GM concerns the clear definition of the conditions and methods for mapping the object information required from the user’s use-case viewpoint for city-scale mapping. The benefits of the B2GM approach are that the user is able directly control the BIM-GIS linkage and integration process in order to acquire the necessary objection information. This can reveal cases of possibly unclear BIM-GIS integration outside the black box in an explicit and standard way, so that the user can distinctively predict the final output obtainable from the BIM-GIS integration. This study examined B2GM in terms of its development background, components, and several utilization examples, as well as the levels and considerations of the integration of different BIM-GIS models.

Building information modeling (BIM) has become mandatory of late in the public sector construction orders, with its utilization in building analysis and modeling increasing. Leadership in Energy & Environmental Design (LEED) is an eco-friendly building certification system managed by U.S. Green Building Council (USGBC). LEED-certified projects vary in their use and evaluation of BIM data according to the building evaluation purpose and use case. This study proposes and implements rule-based LEED evaluation method with BIM (RLEM-BIM), which supports the LEED evaluation automation and variance in connection with BIM from the viewpoint of use cases for the confirmation of its effects. Using RLEM-BIM, this study seeks to reuse the evaluation procedure for LEED project, and to lower the rework rate. To check the effects of using RLEM-BIM, the effects were analyzed through interviews of experts. The proposed evaluation method improves the variance and reuse of the evaluation procedure compared to the existing manual evaluation method.

Augmented reality (AR) systems are becoming next-generation technologies to intelligently visualize the real world in 3D. This research proposes a sensor fusion based pipeline inspection and retrofitting for the AR system, which can be used in pipeline inspection and retrofitting processes in industrial plants. The proposed methodology utilizes a prebuilt 3D point cloud data of the environment, real-time Light Detection and Ranging (LiDAR) scan and image sequence from the camera. First, we estimate the current pose of the sensors platform by matching the LiDAR scan and the prebuilt point cloud data from the current pose prebuilt point cloud data augmented on to the camera image by utilizing the LiDAR and camera calibration parameters. Next, based on the user selection in the augmented view, geometric parameters of a pipe are estimated. In addition to pipe parameter estimation, retrofitting in the existing plant using augmented scene are illustrated. Finally, step-by-step procedure of the proposed method was experimentally verified at a water treatment plant. Result shows that the integration of AR with building information modelling (BIM) greatly benefits the post-occupancy evaluation process or pre-retrofitting and renovation process for identifying, evaluating, and updating the geometric specifications of a construction environment.

The number of scan-to-BIM projects that convert scanned data into Building Information Modeling (BIM) for facility management applications in the Mechanical, Electrical and Plumbing (MEP) fields has been increasing. This conversion features an application purpose-oriented process, so the Scan-to-BIM work parameters to be applied vary in each project. Inevitably, a modeler manually adjusts the BIM modeling parameters according to the application purpose, and repeats the Scan-to-BIM process until the desired result is achieved. This repetitive manual process has adverse consequences for project productivity and quality. If the Scan-to-BIM process can be formalized based on predefined rules, the repetitive process in various cases can be automated by re-adjusting only the parameters. In addition, the predefined rule-based Scan-to-BIM pipeline can be stored and reused as a library. This study proposes a rule-based Scan-to-BIM Mapping Pipeline to support application-oriented Scan-to-BIM process automation, variability and reusability. The application target of the proposed pipeline method is the plumbing system that occupies a large number of MEPs. The proposed method was implemented using an automatic generation algorithm, and its effectiveness was verified.

Accurate earthwork quantity estimation is essential for effective project planning and cost management in the Architecture, Engineering, and Construction (AEC) industry. Traditional methods for quantity takeoff are often time-consuming and susceptible to human error, particularly when working with unstructured datasets such as CAD drawings. This study introduces the Earthwork Network Architecture (ENA), a novel deep learning framework that incorporates Large Language Models (LLMs), Multi-Layer Perceptron (MLP), Long Short-Term Memory (LSTM) networks, and Transformers to automate and enhance the accuracy of earthwork quantity estimation. We assume that if LLMs can be trained effectively using such unstructured construction dataset, the effects such as improved accuracy and the challenges of LLMs can be clearly examined. Among the architectures tested, the LLM-based ENA demonstrated superior performance, achieving faster convergence, greater loss reduction, and higher classification accuracy, with a Quantity Takeoff Classification accuracy of 97.17%. However, the LLMs required significantly more computational resources compared with other models. These findings suggest that LLMs, typically used in natural language processing, can be effectively adapted for complex AEC datasets. This study lays the groundwork for future AI-driven solutions in the AEC industry, underscoring the potential of LLMs and Transformers to automate the quantity takeoff process and manage multimodal data in construction projects.