Article

Intelligent monitoring and evaluation for the prefabricated construction schedule

Authors:
To read the full-text of this research, you can request a copy directly from the authors.

Abstract

The prefabricated building construction (PBC) project is sensitive to uncertainties due to the highly required coordination and interdependency among the installation activities, which may cause progress delay. Hence, it is necessary to monitor the installation progress and evaluate the schedule in terms of the project duration to take proactive control actions to avoid actual project delay. This study focuses on intelligent monitoring and evaluation for the PBC schedule by combining the computer vision‐based (CVB) technology, a weighted kernel density estimation (WKDE) method, and the earned duration management (EDM) method. Intelligent and real‐time far‐field detection of the prefabricated components (PCs) and workers is achieved through the CVB technology, which is, respectively, used to measure the progress status of the PC installation works and other manual works by means of the WKDE method. The PBC project duration is then predicted based on the monitored progress status to evaluate the schedule through the EDM method. The proposed intelligent schedule monitoring and evaluation method have been illustrated and justified through a field application. This study contributes to achieving intelligent schedule monitoring and evaluation for the PBC project.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

... Another study developed a model by which to detect prefabricated walls and slabs, in addition to workers and onsite activities, in order to measure the corresponding progress as a proactive control mechanism for monitoring and evaluating the installation schedule (Yan et al., 2022). Another of the identified studies proposed a model to automatically measure the installation rate of prefabricated panels in order to overcome the limitations of manual methods (which are considered to be time-consuming and highly prone to errors) . ...
... In the model they developed, once the disruption is detected, its impact on the original schedule is evaluated against the set deviation tolerance from the original schedule in order to determine whether the schedule needs to be adjusted to get the project back on track. The other study proposed a computer-visionbased system that measures and evaluates the progress of prefabricated slab/wall installation and other manual work on site, allowing for the detection and timely handling of schedule disruptions (Yan et al., 2022). ...
... It was used in the identified studies to detect various kinds of objects, including cracks and breakages in precast concrete members (S. J. , workers Yan et al., 2022Yan et al., , 2021, machines , cranes , walls (Yan et al., 2022(Yan et al., , 2021, slabs (Yan et al., 2022(Yan et al., , 2021, panels (C. , safety barricades (C. ...
Thesis
Full-text available
The offsite construction industry continues to rely on experience-based average production rates (i.e., average quantity per unit of time) to estimate and schedule production operations. This approach is hindered by various sources of production variability, such as machine breakdowns and material shortages, often resulting in high production estimation and scheduling errors; in fact, as described herein, using average production rates may result in overly optimistic production schedules, leading to missing schedule deadlines, cost overruns, and, most critically, an overburdened workforce. In this context, this thesis proposes a digital twin to enable dynamic production estimation, scheduling, and real-time monitoring of production operations in offsite construction with more accuracy compared to the current practice. The proposed digital twin comprises three major subsystems: (1) an estimation and scheduling subsystem, which estimates variable cycle times as a function of various factors that influence them and virtually mimics operations to estimate production time and generate production schedules; (2) a computer-vision-based data acquisition subsystem that enables the continuous collection of data necessary for regular tuning of the estimation models, accommodating new sources of variability; and (3) a real-time monitoring subsystem to monitor production operations in real time, tracking progress on production schedules and enabling the generation of updated schedules promptly in response to any deviations from the actual operations. To support the development of these subsystems and their requisite functionalities, four main research objectives are pursued: (1) develop and examine a system that deploys computer-vision technology for the automated and accurate acquisition of cycle time data in a timely and cost-effective manner; (2) devise a methodical approach for the identification and understanding of the factors driving cycle time variability, and evaluate how this identification process improves the accuracy of cycle time estimation; (3) design and develop a data- and knowledge-driven system that estimates cycle times in consideration of various influencing factors and using automatically collected data to increase the estimation accuracy compared to traditional estimation methods; and (4) devise a feasible design of a digital twin that enables dynamic and more accurate production estimation, scheduling, and real-time monitoring in offsite construction factories. A diverse array of methods and technologies, including computer vision, 3D simulation, machine-learning-based prediction, statistical modelling, ultrasonic sensors, semi-structured interviews, direct observation, and literature reviews, are deployed and integrated to achieve these objectives. A prototype of the digital twin is developed for a wall framing workstation within a panelized construction factory. The results show that average errors of less than 1 minute in data acquisition, a 36% reduction in cycle time estimation errors, and an 81% reduction in deviations between the production schedule and actual production are achieved compared to the current practice of relying on experience-based average production rates.
... Many studies have utilized the DLBCV techniques to assist in construction management. For example, these techniques have been used to identify unsafe actions [1,2], estimate earthmoving productivity [3], calculate material quantities [4], and evaluate project schedule disruptions [5][6][7]. ...
... With the rapid development of computer hardware and deep learning algorithms becoming a breakthrough in computer vision studies [16], there has been a significant improvement in the speed, accuracy, generalization, and robustness of computer vision algorithms. After being introduced to the construction industry, deep learning-based computer vision (DLBCV) techniques have been widely used by scholars and engineers in construction management; for example, worker posture and behavior identification [1,17], earthmoving productivity estimation [3], material quantity measurement [4], project schedule recovery [5][6][7], overheight vehicle collision monitoring [18], and defect detection [19][20][21]. Meanwhile, security cameras have become widely accepted as monitoring equipment on many construction sites due to their non-invasive approach and ability to capture detailed on-site visual scenes [22]. ...
... Roberts et al. developed an object detection model trained on the advanced infrastructure management group (AIM) dataset [50] to analyze earthmoving equipment productivity [26]. Based on object detection of construction workers, PCs, and trucks, Yan et al. proposed disruption monitoring and evaluation methods for prefabricated building construction (PBC) projects [5][6][7]. Semantic segmentation provides inference by predicting the object category for every pixel in the image and annotating each pixel according to the object category within which it is enclosed [27]. For example, Wang et al. collected and annotated 895 images for 12 object categories in construction activities, based on which a deep semantic segmentation method was trained for visual understanding on construction sites [28]. ...
Article
Deep learning-based computer vision (DLBCV) techniques have played an important role in intelligent construction. Image datasets are essential for developing DLBCV algorithms. However, a large-scale construction-specific dataset of major construction elements, such as precast components (PCs), PC delivery trucks, and workers not wearing safety helmets, remains absent. This paper presents the Construction Instance Segmentation (CIS) dataset, a new image dataset aimed at advancing state-of-the-art instance segmentation in the field of construction management. It contains 50,000 images with ten object categories belonging to construction workers, machines, and materials. Two rounds of algorithmic analysis have been conducted to refine and balance the dataset. Finally, a detailed statistical analysis of the dataset is presented.
... Additionally, computer vision tools have been used to detect precast concrete walls and track their trajectory during the hoisting and installation processes [56,57]. Another study developed a model by which to detect prefabricated walls and slabs, in addition to workers and onsite activities, in order to measure the corresponding progress as a proactive control mechanism for monitoring and evaluating the installation schedule [58]. Another of the identified studies proposed a model to automatically measure the installation rate of prefabricated panels in order to overcome the limitations of manual methods (which are considered to be time-consuming and highly prone to errors) [59]. ...
... In the model they developed, once the disruption is detected, its impact on the original schedule is evaluated against the set deviation tolerance from the original schedule in order to determine whether the schedule needs to be adjusted to get the project back on track. The other study proposed a computer vision-based system that measures and evaluates the progress of prefabricated slab/wall installation and other manual work on site, allowing for the detection and timely handling of schedule disruptions [58]. ...
... The Faster R-CNN algorithm is a deep convolutional network used primarily for near-real-time and accurate object detection [91]. It was used in the identified studies to detect various kinds of objects, including cracks and breakages in precast concrete members [67], workers [58,62,81], machines [62], cranes [62], walls [58,81], slabs [58,81], panels [92], safety barricades [92], and fences [92]. Regarding the performance of the identified applications, it should be first noted that (1) some of these studies evaluated the performance of their overall methodology (which included computer vision tasks in addition to other tasks) as well as the performance of the computer vision algorithms themselves, whereas others reported only the performance of their overall methodology, and (2) some of the studies did not report on the performance either of the developed algorithms or of the overall methodology. ...
Article
The field of computer vision has undergone rapid growth in recent years, yet the use of computer vision in offsite construction remains an under-researched area of study. Given the current momentum around the adoption of this technology, this article presents a scoping review of computer vision applications in offsite construction. It provides (1) summaries of and discussions on the research areas in which computer vision is used in offsite construction, the computer vision tasks undertaken, the algorithms used, and related performance evaluation results and limitations, (2) a tabulated summary of performance-related terms commonly used in computer vision applications (to facilitate understanding of the performance evaluation results reported in the review), and (3) potential avenues of future research. The review provides a useful point of reference for practitioners and researchers in the offsite construction industry, aiding their understanding of current practice, limitations, research gaps, and potential opportunities to apply computer vision.
... Xu proposes an assembly building monitoring method based on feature extraction and point cloud segmentation, which can find the quality problems caused by schedule delays and errors in construction in a timely manner [20]. In the progress, safety, and management of the project, Yan uses computer vision technology, the weighted kernel density estimation method, and the labor duration management method to intelligently monitor and evaluate the progress of prefabricated buildings [21]. Shen uses Autodesk Revit 2016 combined with ontology theory to establish a construction monitoring system for prefabricated components to provide timely information for construction safety risk decision making [22]. ...
... Contract informatization management [56] Contract management informatization (C63) 15. Progress simulation optimization [21] 40. Platform for engineering collaborative management [57] Informatization of an engineering collaborative management platform (C71) 16. ...
Article
Full-text available
Owing to its rapid advancement, information technology has emerged as a critical tool in assembly construction for addressing market demands, improving project quality, and reducing costs. However, the absence of unified informatization standards within the assembly construction industry has led to the adoption of different technologies and systems by various businesses during the development of informatization systems; this has generated issues such as unbalanced development and mutual incompatibility. While researchers have examined these issues, a comprehensive assessment of the maturity of informatization in assembly-building projects is lacking. Assessment of the maturity of informatization can provide evaluation standards and methods for the development of informatization of assembly buildings, explore the important and difficult points of applying informatization technology to assembly buildings, and put forward corresponding countermeasures and suggestions to promote the benign development of informatization of assembly buildings. Therefore, this study strives to develop a model for assessing the maturity of informatization of assembly-building projects. This study begins by determining the level of the maturity level of informatization, key process areas, and key practices for assembly-building projects using the capability maturity model (CMM). On this basis, the maturity evaluation index system was constructed through expert interviews and questionnaires. Furthermore, in order to assign weights to the indicators comprehensively, the ordinal relationship method and entropy weight method were implemented. The evaluation criteria were determined by consulting the relevant literature and expert opinions. Followingly, an evaluation model was established based on the cloud matter element (CME) theory. Finally, a case study demonstrates that the methodology can be utilized to quantify the maturity of project informatization. In conclusion, this study unearths a system for assessing the level of maturity of informatization of assembly-building projects, which provides a valuable reference for promoting the continuous development of the maturity of informatization in assembly-building projects.
... Similarly, Zheng et al. (2020) (Zheng et al., 2020) employed mask R-CNN to detect modules during these stages, but simply detecting the modules does not adequately convey the entirety of the module installation process, especially when interactions between modules and mobile cranes occur. Subsequent studies such as those of Yan et al. (2023) (Yan et al., 2023) and Wang et al. (2021) (Wang et al., 2021) demonstrated the viability of vision-based approaches in OSC, focusing on schedule monitoring and tracking the installation of precast concrete walls. Nonetheless, these studies primarily tracked a limited range of components and did not sufficiently address the sequential operational patterns inherent in the on-site installation of prefabricated components. ...
... Similarly, Zheng et al. (2020) (Zheng et al., 2020) employed mask R-CNN to detect modules during these stages, but simply detecting the modules does not adequately convey the entirety of the module installation process, especially when interactions between modules and mobile cranes occur. Subsequent studies such as those of Yan et al. (2023) (Yan et al., 2023) and Wang et al. (2021) (Wang et al., 2021) demonstrated the viability of vision-based approaches in OSC, focusing on schedule monitoring and tracking the installation of precast concrete walls. Nonetheless, these studies primarily tracked a limited range of components and did not sufficiently address the sequential operational patterns inherent in the on-site installation of prefabricated components. ...
Article
Full-text available
Effectively monitoring and analyzing on-site module installation for modular integrated construction (MiC) is essential to properly coordinating the MiC process. In this study, the authors propose an automated productivity monitoring framework for on-site module installation operations consisting of three modules: object detection, activity classification, and productivity analysis. The object detection module detects mobile cranes and modules interacting with mobile cranes, and the activity classification module classifies module installation activities into five different activities by considering the spatiotemporal relationship between the detected objects. Finally, the productivity analysis module analyzes the productivity of the module installation process by utilizing the accumulated activity classification results over image frames. The proposed model achieves an average accuracy of 89% (hooking: 85.71%, lifting: 84.44%, positioning: 94.90%, returning: 83.09%, and idling: 96.87%) in classifying the five activities. The developed framework enables practitioners to measure the productivity of the on-site module installation process automatically. In addition, productivity data stored from diverse construction sites contribute to identifying progress-impeding factors and improving the productivity of the entire MiC process.
... Rigorous project planning lays the foundation, by setting clear goals, deadlines, and resource allocations (Adebayo, Eniowo & Ogunjobi, 2018). This planning is complemented by robust monitoring mechanisms that track progress against these benchmarks, utilising project management tools to ensure timely and budget-compliant execution (Yan, Zhang & Zhang, 2023). In addition, budget monitoring safeguards against cost overruns, while stringent quality control processes uphold standards and reduce the likelihood of rework (Srewil & Scherer, 2013;Hazır, 2015). ...
... As a result, continuous monitoring is essential to ensure project success and to identify potential issues early [2][3][4][5][6][7][8]. While effective construction process monitoring is critical, conventional manual monitoring methods face limitations in meeting the demands of modern construction sites due to the complexity of projects and limited on-site personnel [3,4,[9][10][11][12]. As projects grow in scale and complexity, the limitations of these manual approaches become increasingly apparent [13][14][15][16]. ...
Article
Full-text available
In the construction industry, ensuring the proper installation, retention, and dismantling of temporary structures, such as jack supports, is critical to maintaining safety and project timelines. However, inconsistencies between on-site data and construction documentation remain a significant challenge. To address this, this study proposes an integrated monitoring framework that combines computer vision-based object detection and document recognition techniques. The system utilizes YOLOv5 for detecting jack supports in both construction drawings and on-site images captured through wearable cameras, while optical character recognition (OCR) and natural language processing (NLP) extract installation and dismantling timelines from work orders. The proposed framework enables continuous monitoring and ensures compliance with retention periods by aligning on-site data with documented requirements. The analysis includes 23 jack supports monitored daily over 28 days under varying environmental conditions, including lighting changes and structural configurations. The results demonstrate that the system achieves an average detection accuracy of 94.1%, effectively identifying discrepancies and reducing misclassifications caused by structural similarities and environmental variations. To further enhance detection reliability, methods such as color differentiation, construction plan overlays, and vertical segmentation were implemented, significantly improving performance. This study validates the effectiveness of integrating visual and textual data sources in dynamic construction environments. The study supports the development of automated monitoring systems by improving accuracy and safety measures while reducing manual intervention, offering practical insights for future construction site management.
... According to the Ovako Working Posture Analysis System (OWAS [59]), the working postures of workers identified during rebar tying activities are ergonomically significant to the productivity and work duration [60]. Rebar tying requires workers to work in a sustained postures [61], and three postures are identified in the rebar tying process: (1) standing, (2) stooping, and (3) squatting. ...
... Wang et al. (2021a, b) also used CCTV videos to detect the walls in the video, projected the BIM model onto the image, and compared the detected walls with the BIM model to estimate the progress of work. Finally, Yan et al. (2023) have recently proposed an intelligent progress monitoring and evaluation method using computer vision to detect the prefabricated walls and slabs in the video and using the Earned Duration Management (Khamooshi and Golafshani 2014) method to evaluate the progress of work against the schedule. ...
Article
Full-text available
Modular and offsite construction methods are being increasingly adopted due to the advantages they offer in terms of project completion time, quality, and energy-efficiency. Despite these advantages, the current state of monitoring systems in modular construction factories highly relies on labor-intensive, subjective, and error-prone observational methods. A large body of research has aimed to automate the monitoring process using an array of sensors, such as IMUs and RFIDs, during the past two decades. Recently, computer vision-based methods have gained increasing interest as a non-intrusive technology to monitor the process inside modular construction factories. However, partial occlusion challenges have impeded their practical application on a large scale. This challenge is specifically important for monitoring the installation of subassemblies since they can obstruct the view of the monitoring camera, especially those that enable long-term monitoring like closed-circuit television (CCTV) fixed-view surveillance cameras. This paper aims to address this challenge by proposing a novel computer vision-based method to monitor the installation of new subassemblies inside modular factories in highly occluded scenes. The proposed methodology identifies the subassemblies in the CCTV video footage using computer vision, analyzes the occlusions using BIM and ray casting techniques, and estimates the progress of assembly by comparing the BIM model with the detected subassemblies in the video. The proposed methodology was successfully validated on surveillance videos captured from a volumetric modular construction factory in the U.S., achieving 93% accuracy in identifying the installation of subassemblies. The results from this research show that the integration of BIM and computer vision is a promising method for monitoring the installation processes inside modular factories under severe occlusion.
... Rigorous project planning lays the foundation, by setting clear goals, deadlines, and resource allocations (Adebayo, Eniowo & Ogunjobi, 2018). This planning is complemented by robust monitoring mechanisms that track progress against these benchmarks, utilising project management tools to ensure timely and budget-compliant execution (Yan, Zhang & Zhang, 2023). In addition, budget monitoring safeguards against cost overruns, while stringent quality control processes uphold standards and reduce the likelihood of rework (Srewil & Scherer, 2013;Hazır, 2015). ...
Article
Full-text available
The construction industry in the Eastern Cape province of South Africa plays a crucial role in the region’s economy, yet small and medium enterprise (SME) construction companies face significant challenges in achieving long-term business sustainability. This article aims to provide practical guidelines for SMEs, by investigating the most critical construction management practices adopted by SME contractors. Utilising a quantitative approach, data were collected from 59 purposefully selected participants, including directors, construction managers, quantity surveyors, site agents, and technicians, all registered under the Construction Industry Development Board (CIDB) Grades 1-4 in General Building (GB). The management practices of SME contractors were analysed using the relative importance index (RII) and factor analysis to rank these practices according to their significance. Findings indicate that SME contractors prioritise health and safety strategies, effective resource utilisation, integrated project management systems, competent recruitment, strong leadership skills, and robust health and safety management as essential practices. The principal component analysis identified six key management factors to enhance competitiveness while contributing to the region’s sustainable development goals. These factors include shared knowledge among management, effective project scope planning, comprehensive health and safety management, ownership of construction business knowledge, scope control on projects, and clearly defined goals for management teams. This study is original in its focus on the specific management practices that can strengthen the sustainability of SME contractors within a developing region, providing a valuable framework for enhancing their operational effectiveness.
... Prefabricated components are standardized products and their quality is relatively high, so there would be no largescale damage in the event of natural disasters such as earthquakes. At the same time, when the prefabricated components are transported to the site for installation after the completion of production in the factory, there would be no large errors or deformation [6]. ...
Chapter
Full-text available
Prefabrication is characterized by standardization of building parts and plant installation. Prefab construction in China is relatively late, but develops quickly. It is a kind of advanced construction method. Along with the development of science and technology, the requirement of building quality is higher, and prefab building has the advantages of high efficiency, low cost, and low pollution, which is consistent with the idea of green development. To guarantee the construction quality of prefab construction works, it is essential to supervise and control the construction process. In this paper, the construction supervision and quality control of prefab construction are analyzed from the angle of digital technique. It is proved by the experiment that the digital technique study on the construction supervision and quality control of the prefab construction project has achieved the highest pass rate of 95.7%.
... For instance, computer vision algorithms can automatically identify safety violations, such as workers not wearing hard hats or harnesses in designated areas, by analyzing video streams from cameras around the site. Additionally, these systems can track the movement of heavy machinery to prevent collisions and ensure that equipment is being used safely and efficiently [28][29][30][31]. Another noteworthy application of computer vision in construction monitoring involves using time-lapse photography and AI analysis to track project progress against planned schedules [32]. ...
Article
Full-text available
In the construction industry, the imperative for visual surveillance mechanisms is underscored by the need for safety monitoring, resources, and progress tracking, especially with the adoption of vision intelligence technology. Traditional camera installation plans often move toward coverage and cost objectives without considering substantial coverage overlap, inflating processing and storage requirements, and complicating subsequent analyses. To address these issues, this research proposes a voxel-based site coverage and overlapping analysis for camera allocation planning in parametric BIM environments, called the PBA approach. The first step is to collect information from the BIM model, which is the input for the parametric modeling step. After that, the PBA approach simulates the virtual devices and the construction layout by employing visual language programming and then generates a coverage area. Lastly, the performance simulation and evaluation of various placement scenarios against predefined criteria are conducted, including visual coverage and overlapping optimization for eliminating data redundancy purposes. The proposed approach is evaluated through its application to construction projects. The results from these various implementations indicate a marked decrease in data overlap and an overall enhancement in surveillance efficacy. This research contributes a novel, BIM-centric solution to visual information adoption in the construction industry, offering a scalable approach to optimize camera placement while mitigating overlapping areas.
... Moreover, they utilized computer vision to detect prefabricated components and workers remotely, and then used the weighted kernel density estimation method to determine the progress status of assembly operations and other operations. Finally, the duration of the AB construction project was predicted based on the monitored schedule status, and the schedule was evaluated by striving for the schedule management method [6]. To show the data of AB intelligent construction, Ouyang and other researchers carried out case design and analysis based on P-ISOMAP algorithm and BIM technology. ...
... To control the construction of PC, the construction process of PC is first divided. It mainly includes four aspects, namely determining the construction plan, producing PC, transporting and storing PC, and assembling PC [16]. Therefore, the construction process includes three stages: assembly, fabrication, and transportation. ...
Article
Full-text available
With the development of prefabricated buildings in China, the demand for prefabricated components is also increasing. The construction schedule of prefabricated components has heterogeneity and timeliness, which makes the traditional scheduling models not applicable. In order to control the construction process and reduce costs, research is conducted on controlling the construction process of prefabricated components in prefabricated buildings. This study divides the construction process into three stages according to the construction characteristics of prefabricated buildings. The scheduling models of these three stages are established, namely assembly, production, and transportation stages scheduling models.. The scheduling model of the three stages are related to each other through the duration constraints. In addition, an improved genetic algorithm is developed to solve the scheduling model of the assembly stage. Then an improved particle swarm optimization is designed to solve the scheduling model in the production and transportation stages. The results show that the minimum duration of the assembly phase was 8 days. The duration and cost of the production phase cannot be minimized at the same time. The minimum carbon emission duration and transportation cost in the transportation phase are 93.8 hours and 22516 yuan, respectively. The improved genetic algorithm tended to flatten out after nearly 180 iterations. The maximum running time of the improved particle swarm algorithm on the training set is 4.23s, the maximum hyper volume is 0.736, and the maximum anti generation distance is 2.35×10 -3 . The scheduling models of different stages and corresponding solving algorithms are effective and provide technical support for the construction process control of assembly parts. The technical contribution of this study is to optimize the genetic algorithm based on weed invasion algorithm and improve the local search ability of genetic algorithm. Then, the differential evolution algorithm is used to improve the particle swarm optimization algorithm and continuously generate new particles to replace the optimal position.
... In addition to traditional Convolutional Neural Networks (CNNs), other models such as Region-CNN and Mask-Region-CNN have found application in monitoring and classifying workers' activities on construction sites [19], [20]. Furthermore, the Region-CNN was modified to detect slabs, walls, and workers for prefabricated building construction [21]. These benefits evaluate worker postures, building conditions, and risk analysis during task execution [22], [23]. ...
... As usual for construction project planning, precast project schedules are sensitive to uncertainties due to the required high level of coordination and interdependences among the activities. It is crucial to assess the schedule in terms of project and activity duration while also keeping track of the progress being made in order to prevent actual project delays [5]. Moreover, precast elements can be located using geospatial tracking technologies, which allow interactions with physical assets through communication with tags or sensors [6]. ...
Article
Purpose The construction and real estate sectors are vital to national economies, but traditional construction methods often lead to challenges such as safety risks, noise and environmental pollution. While intelligent construction is believed to mitigate these issues, there is a lack of solid empirical evidence on whether it truly benefits the general public. This paper seeks to explore the societal benefits of intelligent construction from the public’s perspective, addressing this research gap. Design/methodology/approach The research adopts a two-step approach. First, topic mining is conducted to identify topics closely related to the public’s daily life, such as environmental impact, construction traffic management and construction technologies. These topics are then analyzed through sentiment analysis using a bidirectional long short-term memory model with attention mechanism to determine whether the public has a favorable view of these aspects of intelligent construction, indirectly demonstrating the benefits to the public. Findings The primary topics identified include “industry development,” “technology enterprise,” “construction equipment,” “intelligent technology,” “environmental protection,” “robots” and “construction traffic management.” Sentiment analysis shows that public sentiment is overwhelmingly positive across all topics and regions, with “environmental protection,” “construction traffic management” and “robots” receiving the most favorable reactions. Originality/value This study provides empirical evidence of the societal benefits of intelligent construction from the public’s viewpoint using social media data. The results highlight the need for continued promotion and adoption of intelligent construction due to its positive impact on society.
Article
Smart construction technology offers fresh avenues for advancing the field of civil engineering. It seamlessly integrates across the entire life cycle of civil engineering projects, encompassing planning, design, construction, and maintenance, thereby fundamentally reshaping the landscape of civil engineering development. Nonetheless, it is essential to recognize that, presently, smart construction’s developmental stage remains relatively nascent. Its progression is subject to a myriad of adoption barriers, and the complex dynamics of their interactions remain insufficiently understood. Therefore, this study aims to (1) explore the barriers to the adoption of smart construction; (2) analyze the impact level of each barrier; and the interaction mechanism between the barriers (3) propose effective strategies to promote the development of smart construction. This study commences by identifying 16 major impediments to the adoption of smart construction through a comprehensive synthesis of existing literature and expert interviews. Subsequently, Euclidean similarity analysis is employed to harmonize varying expert assessments. Following this, the Decision-Making Trial and Evaluation Laboratory model is utilized to ascertain the degree of influence associated with each barrier. Further, the Interpretive Structural Model is employed to establish a hierarchical framework that illuminates the interdependencies among these barriers. Additionally, the Matrice d’Impacts Croisés Multiplication Appliqués à un Classement method is invoked to elucidate the roles and statuses of each barrier. Finally, strategies are proposed based on the results of the analysis. This study offers practical strategies for overcoming barriers and driving the adoption of smart construction, filling a critical gap in understanding by identifying key barriers and providing actionable insights, thus significantly advancing the field and empowering stakeholders for successful implementation and dissemination.
Article
Full-text available
Plenty of studies have been conducted on construction productivity currently. However, most of these studies focus on a specific aspect of construction productivity. For example, some studies may examine how precast concrete affects productivity, while others may explore the relationship between Building Information Modeling (BIM) and productivity. Unfortunately, there haven't been any studies that analyze all the factors that affect construction productivity and provide an outlook for the future. Thus, to fill this gap, in this paper, the authors use the Grounded Theory (GT) to code and analyze the literature, applying the principle of open coding and refinement from the original materials. The objective is to classify the factors that affect productivity in the construction industry and objectively reflect the development direction of productivity research within the industry.
Article
Full-text available
Work‐related musculoskeletal disorders pose significant health risks to construction workers, making it essential to monitor their postures and identify physical exposure to mitigate these risks. This study presents a novel framework for real‐time ergonomic risk assessment of workers in construction environments. Specifically, this study develops a lightweight human pose estimation (HPE) model with a residual log‐likelihood estimation head and adopts pose‐tracking technology to enable real‐time recognition of workers’ three‐dimensional (3D) postures. In particular, this study proposes a novel co‐learning method that enables the HPE model to learn two‐dimensional (2D) and 3D features from multi‐dimension datasets simultaneously, substantially enhancing the model's ability to capture 3D postures from 2D images. The proposed framework facilitates real‐time ergonomic risk assessment, reducing potential risks to construction workers and offering promising practical applications.
Article
Construction workers are vulnerable to excessive workloads that may cause ergonomic risks, disturbing the construction schedule. Hence, ergonomic risks and the responding measures for mitigating them, namely ergonomic measures, need to be predicted and assessed so as to help make decisions about management policies. This study focuses on agent-based simulation to predict ergonomic risks and impacts of ergonomic measures on the construction schedule (ERIEMCS) along with the construction process consisting of physical or light tasks. Workers are regarded as the agents and their behaviors in performing tasks and ergonomic measures are modeled by considering different physical capacities due to different workers’ ages. Energy expenditure–based fatigue quantification and the Ovako working posture analyzing system (OWAS) –based quantification of work-related musculoskeletal disorder (WMSD) risk for an individual worker are the basis for quantitative prediction of the fatigue and the WMSD risks of a crew undertaking the construction process through agent-based simulation. An application study based on a prefabricated construction project is presented to demonstrate and justify the proposed method. The results indicate that ergonomic risks can be effectively reduced by adopting ergonomic measures to balance mitigating such risks and maintaining productivity. Prediction of the ERIEMCS through agent-based simulation prior to commencement of work facilitates investigating the effects of aging workers and planning their working schedules using ergonomic measures in combination with schedule management. This study contributes to the body of knowledge on simulation-based methodology for applying ergonomics in construction management, which is especially meaningful given the trend of labor aging and the direction of industrialization in construction.
Article
Full-text available
The assembly process of “pile-column integration” is proposed in this study and applied in the engineering with the characteristics that most of the pile foundations are end-bearing piles, which is conducive to returning to the normal operation of transportation infrastructure in a timely manner. From the perspective of practical application, the bridge structure components, including pile column and cap beam, are reasonably designed and prefabricated according to the requirements of the reconstruction and expansion project of the old bridge. Through non-destructive testing technologies, the concrete strength, cover thickness of reinforcement, and component size of prefabricated components are monitored and tested to evaluate the quality of full-scale prefabricated bridge substructure for “pile-column integration”. The monitoring results showed that the concrete strength monitoring results of prefabricated components by the rebound method are relatively stable. The concrete strength of the prefabricated components was higher than the design concrete strength and their qualified rate was 100%. According to the monitoring of cover thickness of reinforcement, the measured cover thickness of reinforcement in prefabricated components by electromagnetic induction method fell within the allowable range, and their qualified rates were around 90%. The concrete strength and cover thickness of reinforcement for prefabricated components could meet the design requirements. Although the component size of the prefabricated components could be tested by a 3-D point cloud scanning system, the monitoring effect of a relatively smaller component size still needs to be improved. The quality monitoring of full-scale bridge substructures for “pile-column integration” proved the rationality of prefabrication and the feasibility of non-destructive testing technologies, providing references for the application of “pile-column integration”.
Article
Full-text available
The use of cameras for safety monitoring, progress tracking, and site security has grown significantly on construction and civil infrastructure sites over the past decade. Localization of construction resources is a crucial prerequisite for many applications in automated construction management. However, most existing vision‐based methods perform the analysis in the image plane, overlooking the effect of perspective and depth. The manual and labor‐intensive process of traditional calibration techniques, as well as the busy and restrictive construction environment, makes this a challenging task. This study proposes a framework for automatic camera calibration with no manual intervention. The framework utilizes convolutional neural networks for geometrical scene analysis and object detection, which are used to estimate the location of horizon line, vertical vanishing point, as well as objects with known height distributions. This enables automatic estimation of camera parameters and retrieval of scale. The proposed framework is evaluated on images from two major construction projects in Melbourne, Australia. Results show that the proposed method achieves a minimum accuracy of 90% in estimating proximity of points on the ground and can facilitate further development of vision‐based solutions for safety and productivity analysis.
Article
Full-text available
Cables and hangers are critical components of long‐span bridges, tension forces of them are needed to be accurately measured for ensuring the safety of bridges. Traditionally, cable tension forces are measured by attached accelerometers or elastomagnetic (EM) sensors, however, applying these sensors into engineering practice are time‐consuming, labor‐intensive, and highly dangerous. To address these problems, an unmanned aerial vehicle (UAV)‐based noncontact cable force estimation method with computer vision technologies was proposed in this article. Basic concept of the proposed method is to use the UAV‐installed camera for capturing vibration images of cables from a certain distance and cable dynamic properties are extracted by analyzing captured images. It includes two aspects: (a) a line segments detector (LSD) was employed for detecting cable edges from captured video and a line matching algorithm was further proposed for extracting dynamic displacements; (b) the frequency difference of adjacent higher modal frequencies identified from relative displacements of the cable was employed for cable force calculation to avoid the difficulty of extracting fundamental frequency from UAV‐captured video. It should be noted that relative displacement herein refers to the difference between displacements of two points on the same cable. Advantages of the proposed method lie in that the proposed LSD and matching algorithm are more robust than traditional correlation‐based algorithm for calculating dynamic displacements of bridge cables and it does not need to adjust predefined parameters (i.e., subset size in correlation‐based algorithms). In addition, the combination of relative displacement and frequency difference‐based cable force estimation has the capability of enhancing the Fourier spectrum magnitude of bridge cables and reducing the effect of UAV motion on extraction of cable vibration frequencies. The effectiveness and robustness of the proposed method was verified by using an experimental inclined cable and field‐testing data of a long‐span suspension bridge. Results show that calculated cable forces with UAV technology have a good agreement with reference values measured by attached accelerometers and fixed camera, demonstrating correctness and robustness of the proposed method for cable force estimation.
Article
Full-text available
Detecting and locating precast components to confirm and update components’ status information is the key to construction progress monitoring. There are several ways to collect the information of elements in construction sites, such as manual collection, laser scanning, and tag-based methods. But each of these methods has its own limitations. Considering that effective and accurate construction progress monitoring is fundamental to construction management, this paper proposes a novel framework that integrates the latest computer vision methods to realize automatically monitoring construction progress of precast walls, one of the essential components in precast construction. In this framework, object detection, instance segmentation, and multiple-object tracking are combined to collect precast walls’ location and temporal information from the surveillance videos recording the construction phase. Status information identified and collected is stored as a JavaScript object notation (JSON) format and then sent into a corresponding building information model (BIM) to timestamp the wall components. Each method in the framework is evaluated, respectively, and the demonstration on a real project proves the feasibility, convenience, and efficiency of this vision-based framework. The research results prove the proposed framework’s ability to monitor the construction progress of precast walls automatically. Furthermore, the vital information extracted by the proposed framework contributes to serving application scenarios of the cyber-physical system in construction sites.
Article
Full-text available
Computer vision leveraging deep learning has achieved significant success in the last decade. Despite the promising performance of the existing deep vision inspection models, the extent of models’ reliability remains unknown. Structural health monitoring (SHM) is a crucial task for the safety and sustainability of structures, and thus, prediction mistakes can have fatal outcomes. In this paper, we use Bayesian inference for deep vision SHM models where uncertainty can be quantified using the Monte Carlo dropout sampling. Three independent case studies for cracks, local damage identification, and bridge component detection are investigated using Bayesian inference. Aside from better prediction results, the two uncertainty metrics, variations in softmax probability and entropy, are shown to have good correlations with misclassifications. However, modifying the decision or triggering human intervention can be challenging based on raw uncertainty outputs. Therefore, the concept of surrogate models is proposed to develop the models for uncertainty-assisted segmentation and prediction quality tagging. The former refines the segmentation mask and the latter is used to trigger human interventions. The proposed framework can be applied to future deep vision SHM frameworks to incorporate model uncertainty in the inspection processes.
Article
Full-text available
This paper presents a novel dynamic ensemble learning (DEL) algorithm for designing ensemble of neural networks (NNs). DEL algorithm determines the size of ensemble, the number of individual NNs employing a constructive strategy, the number of hidden nodes of individual NNs employing a constructive–pruning strategy, and different training samples for individual NN’s learning. For diversity, negative correlation learning has been introduced and also variation of training samples has been made for individual NNs that provide better learning from the whole training samples. The major benefits of the proposed DEL compared to existing ensemble algorithms are (1) automatic design of ensemble; (2) maintaining accuracy and diversity of NNs at the same time; and (3) minimum number of parameters to be defined by user. DEL algorithm is applied to a set of real-world classification problems such as the cancer, diabetes, heart disease, thyroid, credit card, glass, gene, horse, letter recognition, mushroom, and soybean datasets. It has been confirmed by experimental results that DEL produces dynamic NN ensembles of appropriate architecture and diversity that demonstrate good generalization ability.
Article
Full-text available
This paper presents a novel dynamic ensemble learning (DEL) algorithm for designing ensemble of neural networks (NNs). DEL algorithm determines the size of ensemble, the number of individual NNs employing a constructive strategy, the number of hidden nodes of individual NNs employing a constructive-pruning strategy, and different training samples for individual NN's learning. For diversity, negative correlation learning has been introduced and also variation of training samples has been made for individual NNs that provide better learning from the whole training samples. The major benefits of the proposed DEL compared to existing ensemble algorithms are (1) automatic design of ensemble; (2) maintaining accuracy and diversity of NNs at the same time; and (3) minimum number of parameters to be defined by user. DEL algorithm is applied to a set of real-world classification problems such as the cancer, diabetes, heart disease, thyroid, credit card, glass, gene, horse, letter recognition, mushroom, and soybean datasets. It has been confirmed by experimental results that DEL produces dynamic NN ensembles of appropriate architecture and diversity that demonstrate good generalization ability.
Article
Full-text available
All projects take place within a context of uncertainty. That is especially noticeable in complex hydroelectric power generation projects, which are affected by factors such as the large number of multidisciplinary tasks to be performed in parallel, long execution times, or the risks inherent in various fields like geology, hydrology, and structural, electrical, and mechanical engineering, among others. Such factors often lead to cost overruns and delays in projects of this type. This paper analyzes the efficiency of the Earned Value Management technique and its Earned Schedule extension, as means of forecasting costs and deadlines when applied to complex hydroelectric power production projects. It is worth noting that this analysis was based on simulation models applied to real-life projects. The results showed that cost forecasting becomes very accurate over time, whereas duration forecasting is not reliably accurate.
Article
Full-text available
Machine learning has played an essential role in the past decades, and in lockstep with the main advances in computer technology. Given the massive amount of data generated daily, there is a need for even faster and more effective machine learning algorithms that can provide updated models for real-time applications and on-demand tools. This paper presents FEMa-A Finite Element Machine classifier-for supervised learning problems, where each training sample is the center of a basis function, and the whole training set is modeled as a probabilistic manifold for classification purposes. FEMa has its theoretical basis in the Finite Element Method, which is widely used for numeral analysis in engineering problems. It is shown FEMa is parameterless and has a quadratic complexity for both training and classification phases when basis functions are used that satisfy certain properties. The proposed classifier yields very competitive results when compared with some state-of-the-art supervised pattern recognition techniques.
Article
Full-text available
Despite recent advances in technologies and equipment for automated progress monitoring, most construction companies worldwide do not utilize them for their projects. This can be due to many reasons, such as the high cost of technologies and equipment, need for skilled staff, and lack of sufficient information about the impact of automated progress monitoring on project performance control. The aim of the present research is to investigate the impact of automated progress monitoring on key project performance indicators: time, cost, and quality. This study prepared based on a survey of contracting and engineering consulting firms in North America, Europe, and the Middle East. In the first part of this study, structural equation modeling (SEM) is used to identify the relations of different factors of project progress monitoring (both conventional and automated) with project performance control. In the second part of the study, a benefit analysis is evaluated based on the sixteen (16) journal and international conference papers and also twenty four (24) international construction projects for which automated progress monitoring was employed. The research findings validate the positive impact of real-time, accurate, and cost-effective automated progress monitoring environments and reveal how automated progress monitoring affects construction project success.
Article
Full-text available
The purpose of this paper is to research the concepts of productivity and ergonomics and to analyze the effect/impact of ergonomics on human productivity and moreover on total productivity. Increase in productivity should be one of the primary goals of a business today. Without a high level of productivity, an organization cannot be successful today. Therefore, it is appropriate, that increased attention should be given to productivity improvement. In a comprehensive literature survey more than 50 different techniques of productivity improvement were cataloged. These are classified as a) technology-based, b) employee-based, c) product-based, d) task-based and e) material-based techniques. Ergonomics is a part of the task-based techniques.
Article
Full-text available
This paper addresses how to estimate the final cost of a project and when the estimate becomes accurate. The performance of five forecasting models drawn from literature was evaluated with data from a sample of actual projects. A stability analysis was carried out in order to identify when the forecasts become stable and accurate for the model that emerged as the most accurate.
Article
Full-text available
Timely and accurate monitoring of onsite construction operations can bring an immediate awareness on project specific issues. It provides practitioners with the information they need to easily and quickly make project control decisions. Despite their importance, the current practices are still time-consuming, costly, and prone to errors. To facilitate the process of collecting and analyzing performance data, researchers have focused on devising methods that can semi-automatically or automatically assess ongoing operations both at project level and operation level. A major line of work has particularly focused on developing computer vision techniques that can leverage still images, time-lapse photos and video streams for documenting the work in progress. To this end, this paper extensively reviews these state-of-the-art vision-based construction performance monitoring methods. Based on the level of information perceived and the types of output, these methods are mainly divided into two categories (namely project level: visual monitoring of civil infrastructure or building elements vs. operation level: visual monitoring of construction equipment and workers). The underlying formulations and assumptions used in these methods are discussed in detail. Finally the gaps in knowledge that need to be addressed in future research are identified.
Article
Full-text available
In recent years the Probabilistic Neural Network (PPN) has been used in a large number of applications due to its simplicity and efficiency. PNN assigns the test data to the class with maximum likelihood compared with other classes. Likelihood of the test data to each training data is computed in the pattern layer through a kernel density estimation using a simple Bayesian rule. The kernel is usually a standard probability distribution function such as a Gaussian function. A spread parameter is used as a global parameter which determines the width of the kernel. The Bayesian rule in the pattern layer estimates the conditional probability of each class given an input vector without considering any probable local densities or heterogeneity in the training data. In this paper, an enhanced and generalized PNN (EPNN) is presented using local decision circles (LDCs) to overcome the aforementioned shortcoming and improve its robustness to noise in the data. Local decision circles enable EPNN to incorporate local information and non-homogeneity existing in the training population. The circle has a radius which limits the contribution of the local decision. In the conventional PNN the spread parameter can be optimized for maximum classification accuracy. In the proposed EPNN two parameters, the spread parameter and the radius of local decision circles, are optimized to maximize the performance of the model. Accuracy and robustness of EPNN are compared with PNN using three different benchmark classification problems, iris data, diabetic data, and breast cancer data, and five different ratios of training data to testing data: 90:10, 80:20, 70:30, 60:40, and 50:50. EPNN provided the most accurate results consistently for all ratios. Robustness of PNN and EPNN is investigated using different values of signal to noise ratio (SNR). Accuracy of EPNN is consistently higher than accuracy of PNN at different levels of SNR and for all ratios of training data to testing data.
Article
Full-text available
The intense competition between the Iranian construction companies has led them to take all appropriate measures to decrease the costs as much as possible. Hence, due to the pivotal role of human resources in construction projects cost, a major part of Iranian construction companies seek their profitability and survival in maximizing the productivity of their operatives. Because of the widespread belief among contractors about the low productivity of daily workers and operatives with basic salary, they commit a major part of their projects activities to sub-contractors. Deployment of sub-contractors by construction firms has become largely conventional in country's construction projects. The aims of this paper is defined as determining the factors and grounds affecting sub-contractors productivity and evaluate their overall negative side effects on project productivity via a structured questionnaire. A total of 31 factors selected and were divided into 7 broad categories. The perceptions of companies managers were asked about the level of effect brought about by the mentioned factors and groups upon productivity on a time based criterion. The analysis indicated that the most important grounds affecting sub-contractors productivity in descending order include: Materials/Tools, Construction technology and method, Planning, Supervision system, Reworks, Weather, and Jobsite condition. Project managers should focus on the identified major grounds and relevant factors in order to improve productivity as long as they commit construction activities to sub-contractors.
Article
Full-text available
The ability to effectively communicate progress information and represent as-built and as-planned progress discrepancies are identified as key components for successful project management that allow corrective decisions to be made in a timely manner. However, current formats of reporting e.g., textual progress reports, progress curves, and photographs may not properly and quickly communicate project progress. Current monitoring methods also require manual data collection and extensive data extraction from different construction documents, which distract managers from the important task of decision making. Therefore, to facilitate progress monitoring, this paper proposes visualization of performance metrics that aims to represent progress deviations through superimposition of four-dimensional 4D as-planned model over time-lapsed photographs in single and comprehensive visual imagery. As a part of the developed system, registration of the 4D model with photographs, augmenting photographs, and occlusion removal for progress images are presented. While contextual information is preserved, the as-built photographs are enhanced and augmented with 4D as-planned model in which the performance metrics are visualized. The augmented photographs provide a consistent platform for representing as-planned, as-built, and progress discrepancies information and facilitate communication and reporting processes.
Article
Full-text available
Current monitoring and control methods rely on massive manual work. As a result, control information is expensive to produce, or is generated irregularly. Additionally, the information is only available infrequently — in many cases after the controlled activity was completed — and its quality and integrity are low. The purpose of the research presented in this paper is to improve monitoring and control information, i.e. to offer it on a daily basis, to improve its quality and integrity and to reduce the cost of generating it. To do all this, the Technion APPC researchers started exploring the use of automated data collection (ADC) technologies. We developed models for automated labor and equipment control, materials management and control as well as monitoring safety measures — all described in this paper. Based on the above, the paper discusses development issues and the prospects for commercializing such systems. Materials management and control is currently the most practical application. On the other end of the spectrum is labor control, which currently is the most challenging area.
Article
Bucket fill factor is of paramount importance in measuring the productivity of construction vehicles, which is the percentage of materials loaded in the bucket within one scooping. Additionally, the locational information of the bucket is also indispensable for scooping trajectory planning. Some research has been conducted to measure it via state-of-the-art computer vision approaches, but their robustness against various environment conditions is not considered. The aim of this study is to fill this gap and six distinctive environment settings are included. Images captured by a stereo camera are used to generate point clouds before being structured into 3D maps. This novel preprocessing pipeline for deep learning is originally proposed and its feasibility has been validated through this study. Moreover, multitask learning is employed to exploit the positive relationship among two tasks: fill factor prediction and bucket detection. Therefore, after preprocessing, 3D maps are forwarded to a faster region with convolutional neu-ral network incorporated with an improved residual neural network. The value of fill factor is acquired via a classification and probabilistic-based approach, which is novel, achieving an inspiring result (overall volume estimation accuracy: 95.23% and detection precision: 92.62%) at the same time.
Article
Monitoring large structures using a vision‐based target‐tracking (TT) system while maintaining the full resolution of high‐speed cameras may limit the data size or the sampling rate selection. Also, similar to wireless sensors networks where data loss often occurs during data transmission, TT signals could possibly lose data due to overexposure. The overall goal of this paper is to demonstrate the validity of compressive sensing (CS) for TT time signal processing when faced with challenges such as data loss. The first part of the study is concerned with signal length where TT signals could be further compressed while original data length is already minimum. Next, CS is investigated for improving and recovering TT signals from the possibility of signal loss as well as enhancing sampling rate for system identification purposes. Two case studies were used to obtain the signals for CS processing. The first case included four field‐monitoring tests of a footbridge that were conducted using different sampling rates TT with limited data length to use full cameras resolution. The second case study was concerned with measuring the shifting of the modal properties of a large‐scale bridge model from white‐noise excitation after earthquake loading. The results show that with limited signal length, lower sampling rates, or data loss, CS techniques can successfully improve TT signals.
Article
Modular integrated construction is one of the most advanced off-site construction technologies and involves the repetitive process of installing prefabricated prefinished volumetric modules. Automatic detection of location and movement of modules should facilitate progress monitoring and safety management. However, automatic module detection has not been implemented previously. Hence, virtual prototyping and transfer-learning techniques were combined in this study to develop a module-detection model based on mask regions with convolutional neural network (Mask R-CNN). The developed model was trained with datasets comprising both virtual and real images, and it was applied to two modular construction projects for automatic progress monitoring. The results indicate the effectiveness of the developed model in module detection. The proposed method using virtual prototyping and transfer learning not only facilitates the development of automation in modular construction, but also provides a new approach for deep learning in the construction industry.
Article
Detecting safety helmet wearing in surveillance videos is an essential task for safety management, compliance with regulations, and reducing the death rate from construction industry accidents. However, it is much challenged by some factors like interocclusion, scale variances, perspective distortion, small object detection, and the carrier recognition of safety helmet. Traditional image‐based methods suffer from them. This article proposes a new methodology for detecting safety helmet wearing, which makes use of convolutional neural network‐based face detection and bounding‐box regression for safety helmet detection. On the one hand, the method can help to recognize the carrier of the safety helmet and detect a multiscale and small safety helmet. On the other hand, deep transfer learning based on DenseNet is introduced and applied using two different strategies, namely, object feature extractor and fine‐tuning for safety helmet recognition. To further improve the recognition accuracy, the network model with two peer DenseNet networks is trained by mutual distillation. Extensive analysis and experiments show that the novel methodology has considerable advantages in detecting safety helmet wearing compared to other state‐of‐the‐art models. The proposed model has achieved 96.2% recall, 96.2% precision, and 94.47% average detection accuracy. These results, precision‐recall (PR) curve, and receiver operating characteristic (ROC) curve demonstrate the feasibility of the new model.
Article
Using drones to monitor construction site is a flexible approach with low manpower costs. We study a drone scheduling problem in which a drone flies on a fixed path over a construction site. The flying speed of the drone is optimized to ensure that the drone spends the most time monitoring areas of the construction site that deserve the most attention while ensuring that the drone completes the route within a certain time and without depleting its battery. We propose a nonlinear optimization model for the problem and develop a dynamic programming algorithm to solve the model. A case study is carried out to demonstrate the applicability of the proposed drone scheduling model.
Article
Struck‐by accidents often cause serious injuries in construction. Monitoring of the struck‐by hazards in terms of spatial relationship between a worker and a heavy vehicle is crucial to prevent such accidents. The computer vision‐based technique has been put forward for monitoring the struck‐by hazards but there exists shortages such as spatial relationship distortion due to two‐dimensional (2D) image pixels‐based estimation and self‐occlusion of heavy vehicles. This study is aimed to address these problems, including the detection of workers and heavy vehicles, three‐dimensional (3D) bounding box reconstruction for the detected objects, depth and range estimation in the monocular 2D vision, and 3D spatial relationship recognition. A series of experiments were conducted to evaluate the proposed method. The proposed method is expected to estimate 3D spatial relationship between construction worker and heavy vehicle in a real‐time and view‐invariant manner, thus enhancing struck‐by hazards monitoring at the construction site.
Article
This paper aims at providing researchers and engineering professionals from the first step of solution development to the last step of solution deployment with a practical and comprehensive deep‐learning‐based solution for detecting construction vehicles. This paper places particular focus on the often‐ignored last step of deployment. Our first phase of solution development involved data preparation, model selection, model training, and model validation. Given the necessarily small‐scale nature of construction vehicle image datasets, we propose as detection model an improved version of the single shot detector MobileNet, which is suitable for embedded devices. Our study's second phase comprised model optimization, application‐specific embedded system selection, economic analysis, and field implementation. Several embedded devices were proposed and compared. Results including a consistent above 90% mean average precision confirm the superior real‐time performance of our proposed solutions. Finally, the practical field implementation of our proposed solutions was investigated. This study validates the practicality of deep‐learning‐based object detection solutions for construction scenarios. Moreover, the detailed information provided by the current study can be employed for several purposes such as safety monitoring, productivity assessments, and managerial decision making.
Article
Since project control involves taking decisions that affect the future, the ability to accurately forecast the final duration and cost of projects is of major importance. In this paper, we focus on improving the accuracy of project duration forecasting by introducing a forecasting approach for Earned Value Management (EVM) and Earned Duration Management (EDM) that combines the schedule performance and schedule adherence of the project in progress. As the schedule adherence has not yet been defined formally for EDM, we extend the EVM-based measure of schedule adherence, the p-factor, to EDM and refer to this measure as the c-factor. Moreover, we aim to improve the ability to indicate the expected forecasting accuracy for a project by extending the EVM concept of project regularity to EDM. The introduced forecasting approach and the EDM project regularity indicator are applied to a large number of real-life projects, mainly situated in the construction sector. The conducted empirical experiment shows that the project duration forecasting accuracy can be increased by focusing on both the schedule performance and schedule adherence. Further, this study shows that the EDM project regularity indicator is indeed a more reliable indicator of forecasting accuracy.
Article
Manual construction tasks are physically demanding, requiring prolonged awkward postures that can cause pain and injury. Person posture recognition (PPR) is essential in postural ergonomic hazard assessment. This paper proposed an ergonomic posture recognition method using 3D view-invariant features from a single 2D camera that is non-intrusive and widely installed on construction sites. Based on the detected 2D skeletons, view-invariant relative 3D joint position (R3DJP) and joint angle are extracted as classification features by employing a multi-stage convolutional nerual network (CNN) architecture, so that the learned classifier is not sensitive to camera viewpoints. Three posture classifiers regarding arms, back, and legs are trained, so that they can be simultaneously classified in one video frame. The posture recognition accuracies of three body parts are 98.6%, 99.5%, 99.8%, respectively. For generalization ability, the relevant accuracies are 94.9%, 93.9%, 94.6%, respectively. Both the classification accuracy and generalization ability of the method outperform previous vision-based methods in construction. The proposed method enables reliable and accurate postural ergonomic assessment for improving construction workers' safety and healthy.
Book
Although there has been a surge of interest in density estimation in recent years, much of the published research has been concerned with purely technical matters with insufficient emphasis given to the technique’s practical value. Furthermore, the subject has been rather inaccessible to the general statistician. The account presented in this book places emphasis on topics of methodological importance, in the hope that this will facilitate broader practical application of density estimation and also encourage research into relevant theoretical work. The book also provides an introduction to the subject for those with general interests in statistics. The important role of density estimation as a graphical technique is reflected by the inclusion of more than 50 graphs and figures throughout the text. Several contexts in which density estimation can be used are discussed, including the exploration and presentation of data, nonparametric discriminant analysis, cluster analysis, simulation and the bootstrap, bump hunting, projection pursuit, and the estimation of hazard rates and other quantities that depend on the density. This book includes general survey of methods available for density estimation. The Kernel method, both for univariate and multivariate data, is discussed in detail, with particular emphasis on ways of deciding how much to smooth and on computation aspects. Attention is also given to adaptive methods, which smooth to a greater degree in the tails of the distribution, and to methods based on the idea of penalized likelihood.
Article
Surveillance system is becoming an indispensable system on construction sites with the fast development of computer vision techniques, thus an optimal placement of surveillance cameras is essential for the successful performance of this system. However, to develop effective models and solutions for large‐scale camera placement still remain as opening challenges. Therefore, this study investigated two fundamental placement problems and proposed a multiobjective placement problem, where the maximum‐coverage problem is to monitor the construction layout as much as possible with a limited budget; the minimum‐cost problem is to minimize the cost given a layout required to be fully covered; and the multiobjective problem is to identify the Pareto fronts of cost and coverage ratio of the system. To solve these problems, the objective space and search space were discretized, and the deterministic and heuristic approaches were revised and developed to provide effective solutions. Finally, experiments in a practical project in Hong Kong were conducted to verify the sufficiency of the developed algorithms and findings revealed potential implementations in many scenarios.
Article
Outdoor tasks operated by construction workers are physically demanding, requiring awkward postures leading to pain, injury, accident, or permanent disability. Ergonomic posture recognition (EPR) technique could be a novel solution for ergonomic hazard monitoring and assessment, yet non-intrusiveness and applicability in complex outdoor environment are always critical considerations for device selection in construction site. Thus, we choose RGB camera to capture skeleton motions, which is non-intrusive for workers compared with wearable sensors. It is also stable and widely used in an outdoor construction site considering various light conditions and complex working areas. This study aims to develop an ergonomic posture recognition technique based on 2D ordinary camera for construction hazard prevention through view-invariant features in 2D skeleton motion. Based on captured 2D skeleton motion samples in the test-run, view-invariant features as classifier inputs were extracted to ensure the learned classifier not sensitive to various camera viewpoints and distances to a worker. Three posture classifiers regarding human back, arms, and legs were employed to ensure three postures to be recognized simultaneously in one video frame. The average accuracies of three classifiers in 5-fold cross validation were as high as 95.0%, 96.5%, and 97.6%, respectively, and the overall accuracies tested by three new activities regarding ergonomic assessment scores captured from different camera heights and viewpoints were 89.2%, 88.3%, and 87.6%, respectively. The developed EPR-aided construction accident auto-prevention technique demonstrated robust accuracy to support on-site postural ergonomic assessment for construction workers’ safety and health assurance.
Conference Paper
State-of-the-art object detection networks depend on region proposal algorithms to hypothesize object locations. Advances like SPPnet [7] and Fast R-CNN [5] have reduced the running time of these detection networks, exposing region pro-posal computation as a bottleneck. In this work, we introduce a Region Proposal Network (RPN) that shares full-image convolutional features with the detection network, thus enabling nearly cost-free region proposals. An RPN is a fully-convolutional network that simultaneously predicts object bounds and objectness scores at each position. RPNs are trained end-to-end to generate high-quality region proposals, which are used by Fast R-CNN for detection. With a simple alternating optimization, RPN and Fast R-CNN can be trained to share convolu-tional features. For the very deep VGG-16 model [18], our detection system has a frame rate of 5fps (including all steps) on a GPU, while achieving state-of-the-art object detection accuracy on PASCAL VOC 2007 (73.2% mAP) and 2012 (70.4% mAP) using 300 proposals per image. The code will be released.
Article
The construction industry has a poor productivity record, which was predominantly ascribed to inadequate monitoring of how a project is progressing at any given time. Most available approaches do not offer key stakeholders a shared understanding of project performance in real-time, which as a result fail to identify any project slippage on the original schedule. This paper reports on the development of a novel automatic system for monitoring, updating and controlling construction site activities in real-time. The proposed system seeks to harness advances in close-range photogrammetry to deliver an original approach that is capable of continuous monitoring of construction activities, with progress status determined, at any given time, throughout the construction lifecycle. The proposed approach has the potential to identify any deviation of as planned construction schedules, so prompt action can be taken because of an automatic notification system, which informs decision-makers via emails and SMS. This system was rigorously tested in a real-life case study of an in-progress construction site. The findings revealed that the proposed system achieved a significant high level of accuracy and automation, and was relatively cheap and easier to operate.
Article
The keys for the development of an effective classification algorithm are: 1) discovering feature spaces with large margins between clusters and close proximity of the classmates and 2) discovering the smallest number of the features to perform accurate classification. In this paper, a new supervised classification algorithm, called neural dynamic classification (NDC), is presented with the goal of: 1) discovering the most effective feature spaces and 2) finding the optimum number of features required for accurate classification using the patented robust neural dynamic optimization model of Adeli and Park. The new classification algorithm is compared with the probabilistic neural network (PNN), enhanced PNN (EPNN), and support vector machine using two sets of classification problems. The first set consists of five standard benchmark problems. The second set is a large benchmark problem called Mixed National Institute of Standards and Technology database of handwritten digits. In general, NDC yields the most accurate classification results followed by EPNN. A beauty of the new algorithm is the smoothness of convergence curves which is an indication of robustness and good performance of the algorithm. The main aim is to maximize the prediction accuracy.
Article
With its generally recognized benefits of clean and safe working environment and good quality, prefabricated house construction (PHC) as a solution is gaining momentum in the face of various housing challenges in Hong Kong’s construction industry. Although prefabrication has its own benefits, its fundamental disadvantages of fragmentation, discontinuity, poor interoperability, and scarce real-time information availability have imposed significant adverse influence on the schedule performance of prefabricated house construction. As a result, despite the promise of the government to provide sufficient houses and harmonious housing, schedule delay problems still frequently beset the industry of PHC. To help address schedule delay problems encountered in the construction of prefabrication housing, this research first identified and analysed critical schedule risk factors that may have significant influence on the schedule performance of PHC. Based on the identified schedule risks, the challenges and corresponding required functions for enhancing schedule performance are determined. Then, a radio frequency identification device (RFID)-enabled BIM platform that integrates various involved stakeholders, information/data flow, offshore prefabrication procedures, and state-of-the-art construction technologies, is developed to handle the critical schedule factors. Smart construction objects and RFID-enabled smart gateway work collaboratively to ease operations within the three echelons of prefabrication manufacturing, logistics and on-site assembly construction, while real-time captured data are used to form a closed-loop visibility and traceability mode in which different end users can supervise the construction statuses, progresses in real time. The developed platform can provide various services, tools and mechanisms to different stakeholders, improve the success of daily operations and decision makings throughout PHC management, such that critical schedule risks can be mitigated and the schedule performance of PHC can be enhanced to ensure timely project delivery.
Article
This paper presents an automated system for progress monitoring of key repetitive construction activities in linear infrastructure projects. The prototype model utilizes several feature recognition techniques to automatically measure the actual progress of different locations in the project using high resolution satellite remote sensing images. A spatiotemporal database integrated with a baseline schedule automatically provides the location-based progress data. The system uses location-based charts and a progress map to generate and display the progress reports on a web-based open source interface. The prototype model was applied to an elevated railway project. The main benefits of the proposed system include the automated measurement, enhanced visualization and documentation of planned/actual progress in large linear projects. The system can enhance the process of communication on deviations and construction efficiency between project decision makers through a web-based monitoring interface.
Article
Deeper neural networks are more difficult to train. We present a residual learning framework to ease the training of networks that are substantially deeper than those used previously. We explicitly reformulate the layers as learning residual functions with reference to the layer inputs, instead of learning unreferenced functions. We provide comprehensive empirical evidence showing that these residual networks are easier to optimize, and can gain accuracy from considerably increased depth. On the ImageNet dataset we evaluate residual nets with a depth of up to 152 layers---8x deeper than VGG nets but still having lower complexity. An ensemble of these residual nets achieves 3.57% error on the ImageNet test set. This result won the 1st place on the ILSVRC 2015 classification task. We also present analysis on CIFAR-10 with 100 and 1000 layers. The depth of representations is of central importance for many visual recognition tasks. Solely due to our extremely deep representations, we obtain a 28% relative improvement on the COCO object detection dataset. Deep residual nets are foundations of our submissions to ILSVRC & COCO 2015 competitions, where we also won the 1st places on the tasks of ImageNet detection, ImageNet localization, COCO detection, and COCO segmentation.
Article
State-of-the-art object detection networks depend on region proposal algorithms to hypothesize object locations. Advances like SPPnet and Fast R-CNN have reduced the running time of these detection networks, exposing region proposal computation as a bottleneck. In this work, we introduce a Region Proposal Network (RPN) that shares full-image convolutional features with the detection network, thus enabling nearly cost-free region proposals. An RPN is a fully-convolutional network that simultaneously predicts object bounds and objectness scores at each position. RPNs are trained end-to-end to generate high-quality region proposals, which are used by Fast R-CNN for detection. With a simple alternating optimization, RPN and Fast R-CNN can be trained to share convolutional features. For the very deep VGG-16 model, our detection system has a frame rate of 5fps (including all steps) on a GPU, while achieving state-of-the-art object detection accuracy on PASCAL VOC 2007 (73.2% mAP) and 2012 (70.4% mAP) using 300 proposals per image. The code will be released.
Article
In recent years, a variety of novel approaches for fulfilling the important management task of accurately forecasting project duration have been proposed, with many of them based on the earned value management (EVM) methodology. However, these state-of-the-art approaches have often not been adequately tested on a large database, nor has their validity been empirically proven. Therefore, we evaluate the accuracy and timeliness of three promising deterministic techniques and their mutual combinations on a real-life project database. More specifically, two techniques respectively integrate rework and activity sensitivity in EVM time forecasting as extensions, while a third innovatively calculates schedule performance from time-based metrics and is appropriately called earned duration management or EDM(t). The results indicate that all three of the considered techniques are relevant. More concretely, the two EVM extensions exhibit accuracy-enhancing power for different applications, while EDM(t) performs very similar to the best EVM methods and shows potential to improve them.
Article
In this paper, a real-life project database is created, outranking the existing empirical databases from project management literature in both size and diversity. To ensure the quality of the added project data, a database construction and evaluation framework based on the so-called project cards is developed. These project cards incorporate the concepts of dynamic scheduling and introduce two novel evaluation measures for the authenticity of project data. Furthermore, an overview of the constructed database leads to statements on the difference between planned and actual project performance and on the earned value management (EVM) forecasting accuracy. Moreover, the database is publicly available and can thus become the basis for many future studies related to project management, of which a few are suggested in this paper. To further support these studies, the database will continuously be extended utilizing the project cards. Furthermore, the project cards can also serve didactical purposes.
Article
This paper presents a process management framework for multisensory data fusion for the purpose of tracking the progress of construction activity. The developed framework facilitates the required type of data fusion at any given point in the construction progress, reliably and efficiently. Data are acquired from high-frequency automated technologies such as three-dimensional (3D) imaging and ultrawideband (UWB) positioning, in addition to foreman reports, schedule information, and other information sources. The results of validation through a detailed field implementation project show that the developed framework for fusing volumetric, positioning, and project control data can successfully address the challenges associated with fusing multisensory data by tracking activities rather than objects, a feature that offers superior capability, efficiency, and accuracy over the length of the project. Other contributions of this research include the development of fusion processes that are performed at higher levels of data fusion instead of traditional low-level fusion algorithms, thus supporting decision-making processes and a number of automated construction management applications, such as construction progress tracking, earned-value estimation, and schedule updating.
Article
The concept of schedule monitoring and control as one of the most important functions of project and program management has not been fully exploited. One possible explanation could be the dominance of the Earned Value Management System (EVMS, also known as EVM). EVM was originally developed as a cost management and control tool which was extended to track the schedule as well. EVM and its derivatives (e.g. Earned Schedule) use cost as a proxy to measure schedule performance to control the duration of the project. While there is a correlation between schedule, cost, quality, and scope of a project, using cost to control duration has proven to be misleading. In contrast to Earned Value and Earned Schedule, the authors have developed the Earned Duration Management (EDM) in which they have decoupled schedule and cost performance measures and developed a number of indices to measure progress and performance of schedule and cost, as well as the efficacy and efficiency of the plan at any level of the project. These new indices are easy to understand, have wider applications, and can be used by contractors, clients and the scheduling offices to assess and measure schedule performance. The newly developed duration performance measures are all schedule-based and can be used for forecasting the finish date of the project.
Article
Efficient and effective construction progress tracking is critical to construction management. Current manual methods, which are mainly based on foremen daily reports or quantity surveyor reports, are time consuming and/or error prone. Three dimensional (3D) sensing technologies, such as 3D laser scanners (LADARs) and photogrammetry are now being investigated and have shown potential for saving time and cost for recording project 3D status and thus to support some categories of progress tracking. Although laser scanners in particular and D imaging in general are being investigated and used in multiple applications in the construction industry, their full potential has not yet been achieved. The reason may be that commercial software packages are still too complicated and time consuming for processing scanned data. Methods have however been developed for the automated, efficient and effective recognition of project 3D CAD model objects in site laser scans. A novel system is thus described herein that combines D object recognition technology with schedule information into a combined 4D object recognition system with a focus on progress tracking. This system is tested on a comprehensive field database acquired during the construction of the structure of the Engineering V Building at the University of Waterloo. It demonstrates a degree of accuracy for automated structural progress tracking and schedule updating that meets or exceeds typical manual performance.
Article
In a companion paper, an object-oriented (OO) information model was presented for construction scheduling, cost optimization, and change order management (CONSCOM), based on the creation of a domain-specific development framework. The framework architecture is developed using generic software design elements, called patterns, which provide effective low-level solutions for creating, organizing, and maintaining objects. The OO model has been implemented in a prototype software system for management of construction projects, called CONSCOM, using the Microsoft Foundation Class library in Visual C++. CONSCOM is particularly suitable for highway construction change order management. It can be used by the owner as an intelligent decision support system in schedule reviews, progress monitoring, and cost-time trade-off analysis for change order approval. The OO information model for construction scheduling cost management can be integrated into a concurrent engineering model for the architecture, engineering, and construction industry.
Article
This paper presents a new approach that allows automated recognition of three-dimensional (3D) computer-aided design (CAD) objects from 3D site laser scans. This approach provides a robust and efficient means to recognize objects in a scene by integrating planning technologies, such as multidimensional CAD modeling, and field technologies, such as 3D laser scanning. Using such an approach, it would be possible to visualize the 3D status of a project and automate some tasks related to project control. These tasks include 3D progress tracking, productivity tracking, and construction dimensional quality assessment and quality control. This paper provides an overview of the developed approach and demonstrates its performance in object recognition and project 3D status visualization, with data collected from a construction job site.
Article
The authors were motivated to overcome some of the limitations and shortcomings of the existing software systems for management of construction projects. The result is a new generation software system for CONstruction Scheduling, Cost Optimization, and Change Order Management, which is called CONSCOM. CONSCOM uses the recently patented Neural Dynamics model of Adeli and Park as its computational engine for construction cost optimization and advanced software engineering and object-oriented programming techniques such as framework and pattern. This paper presents some of its recent and innovative capabilities and features. CONSCOM includes an Integrated Management Environment (IME) as its user interface for the effective control and management of construction projects. An example of a highway construction project is presented to demonstrate the unique modelling capabilities of CONSCOM that cannot be modelled by Critical Path Method (CPM) or CPM-like networks.
Article
This paper presents a mathematical model for resource scheduling considering project scheduling characteristics generally ignored in prior research, including precedence relationships, multiple crew-strategies, and time cost trade-off. Previous resource scheduling formulations have traditionally focused on project duration minimization. The proposed model considers the total project cost minimization. Furthermore, resource leveling and resource-constrained scheduling have traditionally been solved independently. In the new formulation, resource leveling and resource-constrained scheduling are performed simultaneously. The proposed model is solved using the patented neural dynamics model of Adeli and Park. A case study is presented to demonstrate the performance of the method under various resource availability profiles.
Article
Recently, the writers developed a general and powerful mathematical model for scheduling construction projects. An optimization formulation was presented with the goal of minimizing the direct construction cost. The nonlinear optimization problem was solved by the recently patented neural dynamics model of Adeli and Park. In this paper an object-oriented (OO) information model is presented for construction scheduling, cost optimization, and change order management (CONSCOM) based on the new construction scheduling model. The goal is to lay the foundation for a new generation of flexible, powerful, maintainable, and reusable software system for the solution of construction scheduling problems. The model is presented as a domain-specific development framework using the Microsoft Foundation Class library and utilizing the software reuse feature of the framework. The framework reuse architecture is more flexible and powerful than other reuse techniques such as components and patterns. A companion paper presents the implementation of the OO information model in a prototype software system for management of construction projects, called CONSCOM.
Article
Ergonomics has traditionally been used to decrease the number of occupational injuries by discovering those postures and tasks that create significant musculoskeletal stresses. However, the principles which underlie ergonomics can potentially be used to improve productivity as well. Ergonomic guidelines may allow prediction of those postures and workplace layouts that maximize the speed at which employees can work. In this study, fifteen subjects performed a typical industrial task in a variety of layouts designed within an ergonomically acceptable work envelope. The effects of tool mass, work height, and movement distance on performance time were measured. All three variables had significant effects on performance time, even within the ergonomic work envelope, however the magnitudes of the effects varied considerably. The results indicate that workstations can be designed to maximize performance and reduce costs by considering both ergonomics and productivity together.
Conference Paper
Proactive replanningattempts to predict scheduling problems or opportunities and adapt to them throughout a schedule's execution. By continuously predicting a task's remaining duration, a proactive replanner is able to accommodate up- coming problems or opportunities before they manifest them- selves. We have developed a kernel density estimation-based method for predicting a task's duration distribution as it ex- ecutes, and have integrated our prediction algorithm with an existing planner based on heuristic repair. Our predictor al- lows the planner to anticipate problems, or opportunities, early enough to avoid, or take advantage of, them, result- ing in executed schedules that score significantly higher on a number of metrics. We have evaluated a limited form of our approach in simulation, and present the results of our ex- periments. Duration prediction achieves an average reward 28.5% higher than the baseline, with 35.9% more reward- laden tasks executed within a fixed horizon.