In the era of the 4th Industrial Revolution, there is a growing trend to integrate new Information Technology (IT) technologies into the construction industry. Among these technologies, smart construction, which utilizes information technology, has ga...
In the era of the 4th Industrial Revolution, there is a growing trend to integrate new Information Technology (IT) technologies into the construction industry. Among these technologies, smart construction, which utilizes information technology, has gained significant attention. Building Information Modeling (BIM) technology, in particular, stands out as a core technology in smart construction and is widely adopted across all sectors of the construction industry. The introduction of BIM technology has enabled automated quantity takeoff and construction cost management through BIM objects, streamlining construction processes.
However, challenges remain when it comes to calculating BIM-based construction costs. There is a lack of established standards for linking the standardized statement with the BIM model, resulting in a dependency on the experience of integration engineers or the preferences of project owners in linking the BIM-derived objects with the statement. Furthermore, while quantity calculations based on BIM are emphasized, there is a significant lack of standardization in the prepared statements.
To address these issues, further efforts are needed to establish robust standards for linking BIM models with standardized statements, ensuring a seamless integration process and accurate cost management in construction projects. By tackling these challenges, the construction industry can fully leverage the potential of BIM technology, improving efficiency, reducing costs, and enhancing overall project outcomes.
Through the existing approaches, the focus of BIM-based construction cost management has primarily been on quantity takeoff, with limited emphasis on standardizing the generated statements. Consequently, there is a need to extract detailed item information from the BIM model and automatically generate standardized Bill of Quantities (BOQ). To address this, the proposed research aims to leverage the benefits of BIM technology and machine learning to create standard BOQs.
The research process involves several steps. First, the BIM model derived from the basic design and the standard detail writing data are collected. The BIM model provides information such as material, volume, area, size, and quantity, while the standard detail data includes construction item, item name, standard, and unit.
Next, the collected data undergoes preprocessing to prepare it for machine learning algorithms. This includes tasks such as removing data redundancy, supplementing incomplete information, and normalizing the data.
Then, a machine learning model is trained using the preprocessed data. Three models, namely Random Forest with Bag-of-Words (RF &BOW), Random Forest with TF-IDF (RF &TF-IDF), and Random Forest with Word2Vec (RF &W2V), are considered for model selection.
Using the trained machine learning model, the research aims to automatically generate standard breakdown data based on the information extracted from the BIM model.
The generated standard BOQ data is then verified by comparing it with existing standard historical data. Any errors or discrepancies found during this verification process are used to refine and correct the prediction model.
Finally, the validated machine learning model is applied to real-world scenarios, completing the process of linking the BIM model with standard specification data.
By implementing this automated workflow, the research aims to enhance the accuracy and efficiency of generating standard BOQs through BIM and machine learning technology. The approach enables better utilization of BIM data, streamlines the standardization process, and improves overall construction cost management.
The preparation of standard Bill of Quantities (BOQ) plays a crucial role in construction projects. However, manual preparation often leads to errors due to human mistakes and consumes significant amounts of time. In contrast, leveraging machine learning algorithms offers the potential to mitigate human errors, streamline the process, and ensure consistent results through automation. By adopting this approach, the time required for preparing standard specifications can be significantly reduced, resulting in faster project completion and cost savings.
The field of automation utilizing machine learning algorithms is continually advancing, offering a range of diverse applications. Through the integration of Building Information Modeling (BIM) and machine learning algorithms, it is anticipated that the automation of standard BOQ creation can be achieved with enhanced accuracy and efficiency. This advancement holds promise for improving the construction industry by enabling more reliable and efficient standard BOQ generation processes.