Posted in | News | Technology

Knowledge Graphs for Safer Construction Sites

A recent review article published in Information explored the application of knowledge graphs (KGs) for automated safety management in construction. The current approaches for creating and using KGs and the related challenges were identified through a scientometric analysis.

Knowledge Graphs for Safer Construction Sites
Study: Knowledge Graphs for Safer Construction Sites. Image Credit: PeopleImages.com - Yuri A/Shutterstock.com

Background

The construction industry remains one of the most perilous sectors. In response, both academic and industry leaders have developed a variety of safety management methods, including performance measurement, policy development, documentation, and risk assessment.

Recent advancements in digital and intelligent platforms are geared towards automated hazard recognition and response. These innovations require enhanced methods for collecting, processing, and analyzing safety data from diverse sources such as sensors, accident reports, and surveillance systems.

Knowledge Graphs (KGs) offer a computing-aided approach to safety management in construction. A KG organizes varied safety-related data into a semantic graph with nodes (or vertices) and edges that represent connections between data from multiple sources. This structure facilitates data retrieval based on semantic relationships among specific data elements.

In the study outlined below, the researchers extensively review the application of KGs in construction safety management and detail a comprehensive process for developing KGs within this domain. The authors utilized the Web of Science and Scopus databases to ensure a robust preliminary analysis, focusing on literature published from 2000 onwards.

Types of KGs

Knowledge Graphs used in safety management can be classified in various ways. Primarily, they are distinguished into static and dynamic KGs based on their ability to incorporate time series data. Traditional KGs represent relationships among nodes without considering temporal changes, which overlooks the dynamic nature of accident occurrences in construction environments.

Dynamic KGs, on the other hand, integrate temporal information either directly into the entities and relations of the graph or by regularly updating the graph's contents. Furthermore, KGs can also be categorized based on their data integration process into single-source and fusion-source KGs. Fusion-source KGs are increasingly popular due to their ability to integrate diverse data sources like videos, event news, meteorological data, and safety regulations.

Additionally, KGs are differentiated as domain-specific and general based on the scope of their entities and edges. In the realm of safety management, domain-specific KGs are preferred for development and analysis over general KGs due to their tailored approach to specific safety concerns.

With the latest technological advancements, KGs are now also being integrated with natural language processing, machine learning, and recommendation systems. These integrations enable KGs to function as databases for complex artificial intelligence (AI) models, including large language models (LLMs), enhancing their applicability and effectiveness in predictive analytics and decision support.

KG Development Process

While there is no standardized procedure for constructing KGs in the computing or construction safety domains, the process typically begins with the integration of internal and external project data. Internal data might include safety files created on-site, while external data can include knowledge not directly related to the project. This external data can expand the knowledge base, refine technical terminology, improve semantic context, and enhance domain-specific knowledge.

The next step involves fitting the external data into predefined ontological templates. These templates act as foundational blueprints for constructing a KG, aimed at reducing information redundancy and simplifying complexity. They facilitate knowledge sharing and ensure semantic interoperability, allowing various stakeholders and computer systems to exchange and retrieve information seamlessly—a core principle of KGs.

Finally, data extraction involves formatting the information into nodes and edges, which are the basic units of a KG. Initial KG structures often consist of two triplets that share a common node, forming node-edge-node relationships. After this, a completion check is performed to identify any issues, such as disconnected segments within the graph. The construction of the KG is considered complete once these problems are rectified, ensuring a robust and interconnected network.

Challenges and Solutions

Several challenges impede the development and widespread adoption of KGs in the construction sector. A primary obstacle is the lack of standardized data formats across various sources, which complicates the establishment of coherent links and hampers the construction of high-quality KGs.

Regulatory documents, which constitute a significant portion of safety management data, often feature complex language structures with safety clauses that lack clear subjects. This complexity undermines the accuracy of automated information extraction and impedes the formation of effective rule triplets essential for developing a clear KG.

Moreover, the specialized jargon prevalent within the construction industry can affect the accuracy of text extraction and contribute to misunderstandings in KG interpretations. Consequently, precise labeling of nodes and relationships is crucial to prevent misinterpretation and ensure the integrity of the graph.

Addressing these issues effectively can involve several strategies. Optimizing data processing through knowledge standardization and employing LLMs, which offer advanced processing capabilities, can significantly enhance KG development. Additionally, generative models for graphs have shown promise in facilitating the evolution of KGs within construction safety management. These models can help mitigate the issues of outdated information and manage the vast datasets typical in construction safety management, thereby simplifying the KG construction process.

Conclusion and Future Prospects

This article provides a thorough review of the development and application of KGs in construction safety management, drawing on insights from 139 relevant publications. The use of KGs extends beyond traditional applications such as risk factor identification, safety recommendations, and risk assessments.

From the analysis of existing limitations, the researchers have proposed three future directions to advance KGs in construction safety management. First, they suggest integrating KGs with graph-based learning algorithms, such as graph neural networks, which could enhance the predictive accuracy and analytical depth of KGs. Second, there is an emphasis on developing dynamic KGs that support real-time safety management, allowing for immediate responses to evolving site conditions. Lastly, establishing a common data environment for construction projects is recommended to standardize data formats and improve interoperability across various project stages.

These strategies are poised to significantly improve the accuracy of KG creation and their effective utilization in enhancing construction safety.

Journal Reference

Kong, F. & Ahn, S. (2024). Use of Knowledge Graphs for Construction Safety Management: A Systematic Literature Review. Information15(7), 390. DOI: 10.3390/info15070390, https://www.mdpi.com/2078-2489/15/7/390

Disclaimer: The views expressed here are those of the author expressed in their private capacity and do not necessarily represent the views of AZoM.com Limited T/A AZoNetwork the owner and operator of this website. This disclaimer forms part of the Terms and conditions of use of this website.

Nidhi Dhull

Written by

Nidhi Dhull

Nidhi Dhull is a freelance scientific writer, editor, and reviewer with a PhD in Physics. Nidhi has an extensive research experience in material sciences. Her research has been mainly focused on biosensing applications of thin films. During her Ph.D., she developed a noninvasive immunosensor for cortisol hormone and a paper-based biosensor for E. coli bacteria. Her works have been published in reputed journals of publishers like Elsevier and Taylor & Francis. She has also made a significant contribution to some pending patents.  

Citations

Please use one of the following formats to cite this article in your essay, paper or report:

  • APA

    Dhull, Nidhi. (2024, July 09). Knowledge Graphs for Safer Construction Sites. AZoBuild. Retrieved on July 24, 2024 from https://www.azobuild.com/news.aspx?newsID=23558.

  • MLA

    Dhull, Nidhi. "Knowledge Graphs for Safer Construction Sites". AZoBuild. 24 July 2024. <https://www.azobuild.com/news.aspx?newsID=23558>.

  • Chicago

    Dhull, Nidhi. "Knowledge Graphs for Safer Construction Sites". AZoBuild. https://www.azobuild.com/news.aspx?newsID=23558. (accessed July 24, 2024).

  • Harvard

    Dhull, Nidhi. 2024. Knowledge Graphs for Safer Construction Sites. AZoBuild, viewed 24 July 2024, https://www.azobuild.com/news.aspx?newsID=23558.

Tell Us What You Think

Do you have a review, update or anything you would like to add to this news story?

Leave your feedback
Your comment type
Submit

While we only use edited and approved content for Azthena answers, it may on occasions provide incorrect responses. Please confirm any data provided with the related suppliers or authors. We do not provide medical advice, if you search for medical information you must always consult a medical professional before acting on any information provided.

Your questions, but not your email details will be shared with OpenAI and retained for 30 days in accordance with their privacy principles.

Please do not ask questions that use sensitive or confidential information.

Read the full Terms & Conditions.