Trustworthy AI and Large Language Models for Construction Safety, Health, and Risk Management

Dear Colleagues,

The construction industry remains one of the most safety-critical sectors, characterized by dynamic operations, uncertain environments, fragmented information, and high-consequence decision making. As artificial intelligence (AI) continues to evolve, large language models (LLMs), multimodal foundation models, computer vision systems, and hybrid knowledge-driven approaches are opening new possibilities for supporting construction safety, health, and risk management. These technologies can assist with hazard identification, safety planning, design-for-safety reviews, regulatory compliance support, training, real-time monitoring, incident investigation and learning, and proactive decision making across the project lifecycle.

However, in safety-critical domains, technical capability alone is not enough. To be practically adopted and responsibly deployed, AI systems must also be trustworthy. This includes transparency, explainability, robustness, reliability, traceability, governance, practicality, and alignment with regulatory and organizational requirements. Construction professionals must be able to understand when and for what tasks AI can be relied upon, how to verify its outputs, and how to integrate it into human-centered workflows without introducing new forms of risk.

This Special Issue seeks high-quality contributions on trustworthy AI and LLMs for construction safety and risk management. We welcome theoretical, methodological, and applied studies that demonstrate how AI can enhance safety performance while addressing key issues such as hallucination, bias, uncertainty, human oversight, domain grounding, and real-world usability. Studies may focus on text-based, multimodal, vision-based, retrieval-augmented, agentic, or hybrid AI systems, as well as their validation in laboratory, simulated, or field contexts.

Topics include, but are not limited to, the following:

• Trustworthy AI frameworks for construction safety and risk management;
• Large language models for hazard identification, safety planning, and safety communication;
• Use of Retrieval-Augmented Generation (RAG);
• Multimodal AI for analyzing drawings, BIM, images, video, documents, and sensor streams;
• Explainable and interpretable AI for safety-critical decision support;
• Human–AI collaboration in safety inspections, pre-task planning, toolbox talks, and design reviews;
• AI-supported prevention through design (PtD) and design-for-safety reviews;
• AI-assisted incident analysis, near-miss learning, and root-cause reasoning;
• Digital twins, BIM, and knowledge graphs for proactive safety risk management;
• Agentic AI for construction workflows, safety documentation, and dynamic risk reasoning;
• Benchmarking, validation, and evaluation protocols for trustworthy AI in construction safety;
• Ethical, legal, organizational, and governance issues in deploying AI for worker safety;
• AI-enabled immersive safety training using VR/AR/XR and conversational agents;
• Wearables, smart sensing, and edge AI for worker health and safety monitoring;
• Low-code/no-code AI for construction safety, health, and risk management.

Dr. S. M. Jamil Uddin
Dr. Ziyu Jin
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 250 words) can be sent to the Editorial Office for assessment.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Buildings is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2600 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

• trustworthy AI
• large language models (LLMs)
• construction safety
• explainable AI
• human–AI collaboration
• multimodal AI
• risk management