Journal of Computing in Civil Engineering

Abstract

Labor shortage and low labor productivity are critical issues in the construction industry. Robotics, as a powerful technology to improve productivity in the digitization era, is being used to assist in addressing these problems. Particularly in recent years, interest in construction robotics research has risen remarkably. To gain a deeper understanding of this burgeoning research field, this study provides a bibliometric analysis of 498 related papers retrieved from WoS (Web of Science). Performance analysis and science mapping were used in the paper to identify the research trends, relationship among authors, top publication sources, regional activities, as well as knowledge base and dominant research sub-fields. The results indicated that the number of papers focusing on construction robotics has been continuously growing since 2013. “Automation,” “system,” and “design” have been the most addressed topics in construction robotics. The findings are indicative of identifying the deficiencies in existing research and provide directions for future research. Those include greater attention paid to the experimental results of on-site construction robots, while ignoring the impacts of practical construction situations, and the development of off-site construction robots needs more support. For theoretical research, there is a lack of studies on human–robot collaboration workflow and management models. Therefore, the study would be valuable in providing practitioners and researchers with a better perspective on the development of construction robotics and facilitating the building of the intellectual wealth of robotics in the construction industry.

Practical Applications

Our study analyzed 498 papers related to construction robotics from 1974 to 2022 through bibliometric approaches (mathematical and statistical methods). The study counted the numbers of construction robotics papers by year, identified most active scholars and top publication sources, and analyzed the relationship among regions. Most importantly, the paper identified the knowledge base of the research field of construction robotics and reviewed various relevant studies from theoretical and practical application perspectives.

Abstract

Construction materials undergo appearance and textural changes during the construction process. Accurate recognition of these changes is critical for effectively understanding the construction status; however, recognizing the various levels of detailed material conditions is not sufficiently explored. The primary challenge in the detailed recognition of the conditions of the material is the availability of labeled training data. To address this challenge, this study proposes a novel state-of-the-art deep learning model that leverages transfer learning, utilizing the pretrained Inception V3 to transfer knowledge to the limited labeled data set in the construction context. This enables the model to learn meaningful representations from the limited training data, enhancing its ability to accurately classify material conditions. In addition, gray-level co-occurrence matrix (GLCM)–based texture features are extracted from the images to capture the appearance and textural changes in construction materials, which are then concatenated with the transferred convolutional neural network (CNN) features to create a more comprehensive representation of the material conditions. The proposed model achieved an overall classification accuracy of 95% and 71% with limited (208 images) and very small (70 images) data sets, respectively. It outperformed different experimental architectures, including CNN models developed using limited data with and without augmentation, CNN model with data augmentation and transfer learning, separate models using local binary pattern (LBP) and GLCM texture features with super learners trained using augmented limited data. The findings suggest that the proposed model, which combines transfer learning with GLCM-based texture features, is effective in accurately recognizing the conditions of construction materials, even with limited labeled training data. This can contribute to improved construction management and monitoring.

Practical Applications

There are several practical applications of the proposed model combining CNN architecture with GLCM textures in construction industry. The primary applications are in activities such as quality inspection and progress monitoring. By analyzing images of different material conditions, such as concrete, plaster, or masonry walls, the model can be used to automatically detect defects or inconsistencies. This enables construction practitioners to ensure the quality of their structures, detect issues early on, and make informed decisions for maintenance and repair. Additionally, the model can be used to monitor the progress of construction projects by analyzing images to track the status and completion of different building components to estimate delays or cost overruns comparing with the expected material condition at a given time of the schedule. Moreover, because the model does not depend on large training data set, it enables construction managers to develop their project specific data sets and automate material condition detection and monitoring.

Abstract

The recent and significant change in the architecture, engineering, and construction (AEC) industry is the increasing use of building information management (BIM) tools in design and how the generated digital data are consumed. Although designers have primarily published blueprints and model files, multiple parties are now involved in producing different aspects of the data and may be consuming data produced by others. This evolution has created new challenges in sharing and synchronizing information across organizational boundaries. Linked building data is a community effort identified as a potential means of overcoming the barriers to data interoperability. In this paper, we present how the industry can be strengthened using a peer-to-peer network based on the InterPlanetary File System (IPFS) framework to address the typical availability problems of web-based data. The paper particularly presents how Linked Data serialization of the Industry Foundation Classes (IFC) standard can be used in the context of IPFS and benchmarks the performance for the publication of models between IPFS versus HTTP protocol. We show that linked building data—in particular, IFC models converted into Resource Description Format (RDF) graphs according to the ifcOWL ontology—can be implemented using the framework, with initial indications of significant benefits of a peer-to-peer network in terms of performance, offline access, and immutable version history. Two use cases in distributed collaborative environments in the AEC/facility management (FM) sector using evolving multidomain models are used to evaluate the work.

Practical Applications

In a construction project, stakeholders must collaborate and share design data models for efficient and safe completion. Building information published using Linked Data techniques aims to facilitate data sharing between interested parties without a central intermediary. However, the challenges of Linked Data stem from using web protocols, resulting in limited availability, suboptimal performance, broken links, and lost history. This study explores using the InterPlanetary File System (IPFS), a communication protocol for computers to communicate with each other directly without needing an intermediary. It addresses issues related to data availability on the web and data-transfer performance. The study demonstrates the suitability of IPFS in addressing the challenges of sharing building data. The data are available even when the original publisher host is down. In addition, the content-based addressing provides a digital fingerprint that can ensure authenticity and uniqueness, eliminating the data’s duplication. Therefore, it minimizes the storage space consumed by data backups and archives. However, because IPFS addresses and communication principles are incompatible with the Linked Data standard approach, we designed a way so that the construction data saved in IPFS can be accessed transparently as a part of the standard Linked Data network.

The design intent and many meaningful semantics in building information modeling (BIM) models are often implicit, and some explicit semantics are lost during model exchanges. The missing information can be artificially supplemented through a process called semantic enrichment. However, previous research on semantic enrichment has primarily focused on specific tasks, leading to a limited scope of predictions, and lacks comprehensive, seamless approaches. In this study, we aim to infer BIM semantics from fundamental model data, pure object geometries, and organize the predicted results into a graph-based common data environment (CDE) to support intelligent applications. Consequently, we propose a framework of generic semantic enrichment that includes four fundamental tasks in the context of graphs and a process control mechanism to execute a set of tools in a proper sequence. To validate its feasibility, we selected a real-world apartment model and developed six tools to generate the graph-based CDE from its object geometries. Additionally, applications were implemented on the graph-based CDE, such as (1) enriching a pure geometry model from SketchUp to a BIM model in Revit; and (2) easing interoperability problems by reconstructing ArchiCAD models in Revit or downgrading Revit models to earlier versions. The experiment’s success demonstrates the feasibility of constructing BIM graphs from object geometries through semantic enrichment to enable intelligent applications. This study establishes a theoretical foundation for graph semantic enrichment and opens a door to further explore intelligent applications on BIM graphs.

The estimation of material volume in a construction vehicle’s bucket is a crucial prerequisite for automation, as well as for productivity assessment and efficient material transport. Although some studies have been conducted in this field, the accuracy and speed of inference have been suboptimal, and specific implementation strategies have not been proposed. To address these issues, this paper proposes a new approach. The proposed approach has three main components. First, a novel image preprocessing framework based on three-dimensional (3D) grayscale terrain is presented. Second, a semantic mask-level data set is constructed to facilitate future research in this area. Third, a combined neural network and probabilistic approach is proposed to estimate the material volume, with speed and accuracy as metrics. Transfer learning is introduced to improve training efficiency and accuracy. The proposed material volume estimation method is implemented on three different devices, addressing the problem from the development phase to the application phase. The advantages and disadvantages of each device are discussed in depth. The results demonstrate that the proposed approach achieves an impressive average accuracy of 98.20% on all three devices, with real-time or semi–real-time volume estimation feasible on each. In summary, this paper proposes a new approach to estimate the material volume in a construction vehicle’s bucket, addressing issues of accuracy and speed of inference and providing specific implementation strategies. The results demonstrate the effectiveness of the proposed approach, which has potential applications in automation and productivity assessment in the construction industry.

One of the most challenging tasks in pavement management and rehabilitation is to detect and classify different distress types from images collected during field surveys. In this paper, a multilabel convolutional neural network (CNN) model for classifying asphalt distress is proposed. Unlike typical CNN models that classify a single object per image, the proposed model can detect and classify multiple distress types per image, without prior knowledge of the distress location. The model can classify the distress types into four categories: alligator cracking, block cracking, longitudinal/transverse cracking, and pothole. The proposed model was trained and tested on a real data set comprising 42,520 images using different pretrained architectures with various hyperparameter combinations. The results demonstrate the robustness of the proposed model and its potential for crack detection and localization using weakly supervised machine learning methods that can cope with partially labeled data sets.

As an essential sector of the construction industry, bridge construction produces a large amount of carbon dioxide and other greenhouse gases, greatly intensifying the global greenhouse effect. Prefabricated reinforced concrete (RC) T-beams are a vital type of structural element widely used in bridge construction projects such as river-spanning bridges, high-speed railways, and urban expressways. This study proposes a bridge structural optimization framework using building information modeling (BIM) and two-stage metaheuristic searching (MS) in order to minimize the carbon emission of prefabricated RC T-beams. Under the constraints of corresponding design specifications, the first stage of the proposed MS is utilized to optimize the longitudinal steel section size and area of the T-beam, with carbon emission as the optimization objective. The optimal combination of rebar diameters is then obtained in the second stage of the proposed MS. An illustrative example is also provided to demonstrate the performance of the proposed framework. Moreover, the proposed framework is compared with conventional approaches as well as other metaheuristic algorithms to validate the developed optimization approach.