Search results

Images taken from construction site interiors often suffer from low illumination and poor natural colors, which restrict their application for high-level site management purposes. The state-of-the-art low-light image enhancement method provides promising image enhancement results. However, they generally require a longer execution time to complete the enhancement. This study aims to develop a refined image enhancement approach to improve execution efficiency and performance accuracy.

To develop the refined illumination enhancement algorithm named enhanced illumination quality (EIQ), a quadratic expression was first added to the initial illumination map. Subsequently, an adjusted weight matrix was added to improve the smoothness of the illumination map. A coordinated descent optimization algorithm was then applied to minimize the processing time. Gamma correction was also applied to further enhance the illumination map. Finally, a frame comparing and averaging method was used to identify interior site progress.

The proposed refined approach took around 4.36–4.52 s to achieve the expected results while outperforming the current low-light image enhancement method. EIQ demonstrated a lower lightness-order error and provided higher object resolution in enhanced images. EIQ also has a higher structural similarity index and peak-signal-to-noise ratio, which indicated better image reconstruction performance.

The proposed approach provides an alternative to shorten the execution time, improve equalization of the illumination map and provide a better image reconstruction. The approach could be applied to low-light video enhancement tasks and other dark or poor jobsite images for object detection processes.

Measuring onsite productivity has been a substance of debate in the construction industry, mainly due to concerns about accuracy, repeatability and unbiasedness. Such characteristics are central to demonstrate construction speed that can be achieved through adopting new prefabricated systems. Existing productivity measurement methods, however, cannot cost-effectively provide solid and replicable evidence of prefabrication benefits. This research proposes a low-cost automated method for measuring onsite installation productivity of prefabricated systems.

Firstly, the captured ultra-wide footages are undistorted by extracting the curvature contours and performing a developed meta-heuristic algorithm to straighten these contours. Then a preprocessing algorithm is developed that could automatically detect and remove the noises caused by vibrations and movements. Because this study aims to accurately measure the productivity the noise free images are double checked in a specific time window to make sure that even a tiny error, which have not been detected in the previous steps, will not been amplified through the process. In the next step, the existing side view provided by the camera is converted to a top view by using a spatial transformation method. Finally, the processed images are compared with the site drawings in order to detect the construction process over time and report the measured productivity.

The developed algorithms perform nearly real-time productivity computations through exact matching of actual installation process and digital design layout. The accuracy and noninterpretive use of the proposed method is demonstrated in construction of a multistorey cross-laminated timber building.

This study uses footages of an already installed surveillance camera where the camera's features are unknown and then image processing algorithms are deployed to retrieve accurate installation quantities and cycle times. The algorithms are almost generalized and versatile to be adjusted to measure installation productivity of other prefabricated building systems.

Recognising the as-built state of construction elements is crucial for construction progress monitoring. Construction scholars have used computer vision-based algorithms to automate this process. Robust object recognition from indoor site images has been inhibited by technical challenges related to indoor objects, lighting conditions and camera positioning. Compared with traditional machine learning algorithms, one-stage detector deep learning (DL) algorithms can prioritise the inference speed, enable real-time accurate object detection and classification. This study aims to present a DL-based approach to facilitate the as-built state recognition of indoor construction works.

The one-stage DL-based approach was built upon YOLO version 4 (YOLOv4) algorithm using transfer learning with few hyperparameters customised and trained in the Google Colab virtual machine. The process of framing, insulation and drywall installation of indoor partitions was selected as the as-built scenario. For training, images were captured from two indoor sites with publicly available online images.

The DL model reported a best-trained weight with a mean average precision of 92% and an average loss of 0.83. Compared to previous studies, the automation level of this study is high due to the use of fixed time-lapse cameras for data collection and zero manual intervention from the pre-processing algorithms to enhance visual quality of indoor images.

This study extends the application of DL models for recognising as-built state of indoor construction works upon providing training images. Presenting a workflow on training DL models in a virtual machine platform by reducing the computational complexities associated with DL models is also materialised.

The short life cycle replacement of fitout in modern high-rise office buildings represents an under-researched waste problem. This paper aims to quantify the amount of demolition waste from office strip-out including attention to waste streams going to landfill, reuse and recycling.

Quantitative waste data (by weight) were measured from 23 office fitout projects situated in “A” grade office building stock from the Sydney CBD. Waste streams were measured separately for landfill, reuse and recycled materials. Descriptive and clustering statistics are presented and analysed.

From a total of 9,167 tonnes office fitouts demolished, 5,042 tonnes are going to landfill. The main contributor to landfill stream is the mixed waste generated in a fast-track demolition process. This approach partly resulted from the office interiors lacking regularity and easy disassembly. Moreover, considerable variability is observed in the waste per area, the waste streams and the waste compositions. Also, it is noteworthy that the recycled waste stream considerably increases when there exist economically viable conversion facilities, as for metals, hard fills and plasterboards.

The research is focused upon work practices that take place in Australia; therefore, generalisability is limited to situations that have similar characteristics. Future studies are needed to verify and extend the findings of this research.

A key area arising from the research findings is the need to design fitout with recycling and reuse in mind to divert more from landfill. This must explore and incorporate onsite demolition processes to ensure the design is well suited to commercially dominant processes in the overall demolition process, as well as attention to developing economies of scale and viability in re-sale markets for reused items.

Little empirical or quantitative research exists in the area of office fitout waste. This research provides entry to this area via quantifiable data that enables comparison, benchmarking and diagnostic ability that can be used to underpin strategic solutions and measurement of improvements.

Accurate and rapid tracking and counting of building materials are crucial in managing on-site construction processes and evaluating their progress. Such processes are typically conducted by visual inspection, making them time-consuming and error prone. This paper aims to propose a video-based deep-learning approach to the automated detection and counting of building materials.

A framework for accurately counting building materials at indoor construction sites with low light levels was developed using state-of-the-art deep learning methods. An existing object-detection model, the You Only Look Once version 4 (YOLO v4) algorithm, was adapted to achieve rapid convergence and accurate detection of materials and site operatives. Then, DenseNet was deployed to recognise these objects. Finally, a material-counting module based on morphology operations and the Hough transform was applied to automatically count stacks of building materials.

The proposed approach was tested by counting site operatives and stacks of elevated floor tiles in video footage from a real indoor construction site. The proposed YOLO v4 object-detection system provided higher average accuracy within a shorter time than the traditional YOLO v4 approach.

The proposed framework makes it feasible to separately monitor stockpiled, installed and waste materials in low-light construction environments. The improved YOLO v4 detection method is superior to the current YOLO v4 approach and advances the existing object detection algorithm. This framework can potentially reduce the time required to track construction progress and count materials, thereby increasing the efficiency of work-in-progress evaluation. It also exhibits great potential for developing a more reliable system for monitoring construction materials and activities.

Request for information (RFI) documents play a pivotal role in seeking clarifications in construction projects. However, perceived as inevitable “non-value adding” tasks, they harbour risks like schedule delays and increased project costs, underlining the importance of strategic RFI management in construction projects. Despite this, a lack of literature dissecting RFI processes impedes a full understanding of their intricacies and impacts. This study aims to bridge the gap through a comprehensive literature review, delving into RFI intricacies and implications, while emphasising the necessity for strategic RFI management to prevent project risks.

This research study systematically reviews RFI-related papers published between 2000 and 2023. Accordingly, the review discusses key themes related to RFI management, yielding best practices for industry stakeholders and highlighting research directions and gaps in the body of knowledge.

Present RFI management platforms exhibit deficiencies and lack analytics essential for streamlined RFI processing. Complications arise in building information modelling (BIM)-enabled projects due to software disparities and interoperability hurdles. The existing body of knowledge heavily relies on manual content analysis, an impractical approach for the construction industry. The proposed research direction involves automated comprehension of unstructured RFI content using advanced text mining and natural language processing techniques, with the potential to greatly elevate the efficiency of RFI processing.

The study extends the RFI literature by providing novel insights into the problemetisation with the RFI process, offering a holistic understanding and best practices to minimise adverse effects. Additionally, the paper synthesises RFI processes in traditional and BIM-enabled project settings, maps a causal-loop diagram to identify associated issues and summarises approaches for extracting knowledge from the unstructured content of RFIs. The outcomes of this review stand to offer invaluable insights to both industry practitioners and researchers, enabling and promoting the refinement of RFI processes within the construction domain.