Journal of Computing in Civil Engineering

The productivity analysis of cable crane transportation in the construction field is of great significance to improve crane equipment management and reduce operation costs. However, the traditional manual recording method of analyzing cable crane productivity is time-consuming and tedious. The existing vision-based method requires significant amounts of time to collect extensive images at construction sites and does not achieve high-precision detection in complex scenes. Thus, an automated vision-based method for productivity analysis of cable crane transportation is proposed using a new synthetic image approach based on an augmented reality (AR) technique. The unmanned aerial vehicle-based three-dimensional (3D) reconstruction of a crane bucket model is superimposed on a realistic scene using AR to synthesize the images for vision-based model training without manual image acquisition at a construction site. The feature pyramid network and attention module are integrated into Faster region-based convolutional neural network (Faster R-CNN) to enhance the capability of feature extraction for the high-precision detection of a crane bucket and its ID number, which provides the logical basis for calculating productivity. The proposed vision-based productivity analysis method is evaluated on large-scale hydraulic engineering. The results demonstrate that the mean average precision (mAP) of detection performance is 98.01% using the model trained by AR-based synthetic images, which confirms the proposed AR-based synthetic image method could provide a new image generation mode for the construction industry. Additionally, the bias of productivity between the proposed method and ground truth is 0.03%, which confirms the effectiveness and accuracy of the proposed method.

Building information modeling (BIM) has been widely accepted in the industry and extensively used in supporting many construction tasks. In the government sector, the USDOT Federal Highway Administration (FHWA) has implemented building information modeling (BIM) for bridge construction. Hence, state DOTs are now faced with heightened pressure in complying with the FHWA’s Bridge Information Modeling (BrIM) standardization. Although BIM can provide many benefits to DOTs, current BIM-based platforms for bridges are not fully developed to process traditional two-dimensional (2D) bridge drawings for BIM-based computational tasks involving existing bridges, for example cost estimation. Bridges are a critical infrastructure in any nation’s economy, and by law the DOTs are tasked with ensuring that they remain safe for use. To maintain bridges, engineers currently perform periodic inspections, assessing each part of the bridge to identify areas that require maintenance. Maintenance work items are then generated for these areas; these are usually computed traditionally or by systems that still rely heavily on manual inputs. Such processes are time-consuming and cumbersome, and depend on years of bridge technical expertise. To overcome these limitations and improve the accuracy of processes such as generating maintenance work items for bridges, we propose a framework for automatically (1) processing existing 2D bridge drawings for bridges built pre-BIM adoption in the architecture, engineering, and construction (AEC) industry; (2) converting these record drawings into three-dimensional (3D) information models; and (3) converting 3D information models into industry foundation class (IFC) files. The developed 3D models using the proposed framework were compared against developed 3D models using the state-of-the-art method. Experimental results show that the developed framework can be used in developing algorithms that generate 3D models and IFC output files from portable document format (PDF) bridge drawings in a semiautomated fashion. The proposed method uses 3.33% of the time it takes the current state-of-the-art method to generate a 3D model, and the generated models are of comparative quality.

For optimally designing water distribution networks (WDNs), the nondeterministic polynomial hard problem, a novel hybrid model, is introduced with the combined features of evolutionary and swarm intelligence techniques. An evolutionary algorithm with better exploration properties, differential evolution (DE), and the swarm intelligence technique with better exploitation properties, namely the krill herd algorithm (KHA), is considered for this purpose. Because exploration and exploitation are the essential features of the metaheuristic algorithms, the hybrid algorithm with a combination of the DE and KHA features, the DE-KHA, resulted in a balanced search methodology. The results on the application of the proposed model on well-studied benchmark problems have demonstrated its enhanced search behavior, converging faster to the promising results with considerable robustness. Moreover, compared with other competing algorithms reported for optimally designing the WDNs, the DE-KHA outperforms with better computational efficiency. Additionally, considering the few control parameters that have to be calibrated for their optimal values, the computational burden will be less for performing the sensitivity analysis. As a result, considering the solution precision, quick convergence ability, and robustness of DE-KHA, the study suggests the algorithm for efficiently handling real-life case studies.

To fully automate building code compliance checking, regulatory requirements need to be automatically extracted and transformed from building codes. Existing regulatory requirement processing efforts mainly focus on building code requirements in text. A more efficient approach for processing regulatory requirements in other parts of building codes, such as tables or charts, remains to be addressed. The ability to process building code requirements in all parts and formats is necessary for an automated code compliance checking system to achieve full coverage of checkable building code requirements. To address this gap, the authors propose a semiautomated information extraction and transformation method. The proposed method can extract building code requirements in tables and convert extracted information to logic rules. Automated code compliance checking systems can utilize the logic rules. The proposed method includes two main steps: (1) tabular information extraction, and (2) rule generation. The tabular information extraction semiautomatically detects the layout of tables, extracts building code requirements from tables, and transforms extracted information to databases. The rule generation step automatically generates logic rules that can be directly executed by logic reasoners. The rule generation step also provides options for users to further refine the generated rules. The development of the tabular information extraction algorithm takes an iterative approach. An experiment was conducted to develop a tabular information extraction algorithm from a section of existing code. The primary version of the algorithm correctly processed 91.67% of tables in a second sample section. After iterative refinements, the updated tabular information extraction algorithm correctly processed all tables in this section. The rule generation algorithm correctly generated logic rules that successfully represented the applicable building code requirements for a testing building, a convenience store in Texas based on the tabular information.