The science of supporting knowledge-intensive activities

Advanced computing methods and related technologies are changing the way engineers interact with the information infrastructure. Explicit knowledge representation formalisms and new reasoning techniques are no longer the sole territory of computer science. For knowledge-intensive tasks in engineering, a new philosophy and body of knowledge called Engineering Informatics is emerging.

Advanced Engineering Informatics solicits research papers with particular emphases both on 'knowledge' and 'engineering applications'. As an international Journal, original papers typically:

• Report progress in the engineering discipline of applying methods of engineering informatics.
• Have engineering relevance and help provide the scientific base to make engineering decision-making more reliable, spontaneous and creative.
• Contain novel research that demonstrates the science of supporting knowledge-intensive engineering tasks.
• Validate the generality, power and scalability of new methods through vigorous evaluation, preferably both qualitatively and quantitatively.

In addition, the Journal welcomes high quality review articles that summarise, compare, and evaluate methodologies and representations that are proposed for the field of engineering informatics. Similarly, summaries and comparisons of full-scale applications are welcomed, particularly those where scientific shortcomings have hindered success. Typically, such papers have expanded literature reviews and discussion of findings that reflect mastery of the current body of knowledge and propose novel additions to contemporary research.

Papers missing explicit representation and use of knowledge, such as those describing soft computing techniques, mathematical optimization methods, pattern recognition techniques, and numerical computation methods, do not normally qualify for publication in the Journal. Papers must illustrate contributions using examples of automating and supporting knowledge intensive tasks in artifacts-centered engineering fields such as mechanical, manufacturing, architecture, civil, electrical, transportation, environmental, and chemical engineering. Papers that report application of an established method to a new engineering subdomain will qualify only if they convincingly demonstrate noteworthy new power, generality or scalability in comparison with previously reported validation results. Finally, papers that discuss software engineering issues only are not in the scope of this journal.

Special Issue on Foundation Models in Architecture, Engineering, and Construction: Reliability, Generalizability, Interpretability and Trust
提出日: 2026-01-31

The development of multi-modal Foundation Models including Large Language Models (LLMs) and Large Vision Models (LVMs) have accelerated at an unprecedented rate and are quickly adapted in engineering. These large-scale multi-modal models can analyze and generate textual and visual content for many applications in Architecture, Engineering and Construction (AEC). These wide capabilities result in many promising results in those fields that also reveal issues around those new models. Those limitation range from issues like incorrect outputs in form of hallucinations, missing interpretability of the large-scale models and limited reasoning capabilities and generalizability in complex situations. This special issue is investigating those strengths and limitations of Foundation Models in AEC.

This special issue aims to explore methodologies that enhance the trustworthiness, interpretability, and robustness of FMs in these contexts. Examples of research questions this special issue seeks to address include: (1) How can we increase the robustness and reliability of FM in engineering applications? (2) What quantitative and qualitative metrics can accurately assess the quality, clarity, and utility of explanations provided by LLMs for engineering decision-making? (3) How can we design standardized benchmarks and datasets to compare interpretability methods across engineering and construction domains? (4) How can we develop explanation methods that scale with the complexity and size of FMs in engineering applications? (5) How do we integrate FMs into workflows that ensure high performance and quality in engineering contexts? (6) How can LLMs generate causal and counterfactual explanations that reveal underlying reasoning and decision pathways in engineering processes? (7) How can insights from interpretability FMs be used to diagnose errors and improve the robustness of LLMs in engineering workflows?

Addressing these questions is pivotal for advancing the responsible development and deployment of FMs in engineering and construction, ensuring they are more powerful while maintaining transparency, accountability, and alignment with ethical and regulatory standards. Topics of interest include, but are not limited to:

Specifically, it will examine approaches that provide: (1) novel FM approaches to increase robustness in AEC (chain-of-thought, neuro-symbolic reasoning, 3D diffusion models); (2) approaches for consistent and reliable knowledge retrieval (explainability, RAG, knowledge graphs, causality) (3) studies on accountability and compliance (e.g., regulatory requirements, safety standards, and ethical considerations); (3) tools for debugging and model improvement (e.g., error analysis, bias detection, and consistency verification); and (4) user confidence and user empowerment (e.g., user interaction, reasoning and decision pathways).

    Novel FM approaches to enhance trustworthiness and reliability in AEC (chain-of-thought, neuro-symbolic reasoning, multi-view image synthesis)
    Approaches for consistent and reliable retrieval of engineering knowledge (knowledge-augmentation, knowledge graphs, causality)
    Approaches to increase explainability and interpretability of FM in engineering (generative approaches, visualization, diagnosability)
    Studies on accountability and compliance of FM to AEC requirements (e.g., regulatory requirements, safety standards, and ethical considerations)
    Tools for debugging and model improvement in AEC (e.g., error analysis, bias detection, and consistency verification)
    Approaches to increase user confidence and empowerment in AEC use cases (e.g., user interaction, reasoning and decision pathways)
    Use cases studies that applicability of FM under engineering constraints (Constrained Generative Design, Assisted Planning, Construction Robotics, Asset Management)
    Large-scale multi-modal benchmark and training datasets for FM in AEC (documents, images, standards, BIM)

Guest editors:

Prof. Weili Fang
Huazhong University of Science and Technology, Wuhan, China

Prof. Peter Love
Curtin University, Perth, Australia

Prof. Joern Ploennigs
Universität Rostock, Rostock, German

Prof. Jane Matthews
Deakin University, Geelong, Australia

Manuscript submission information:

    Open for Submission: from 01-Jul-2025 to 31-Jan-2026
    Submission Site: Editorial Manager®
    Article Type Name: "VSI: ADVEI_FM in AEC" - please select this item when you submit manuscripts online

All manuscripts will be peer-reviewed. Submissions will be evaluated based on originality, significance, technical quality, and clarity. Once accepted, articles will be posted online immediately and published in a journal regular issue within weeks. Articles will also be simultaneously collected in the online special issue.

For any inquiries about the appropriateness of contribution topics, welcome to contact Leading Guest Editor (mailto: Email Address).
Guide for Authors will be helpful for your future contributions, read more: Guide for authors - Advanced Engineering Informatics
For more information about our Journal, please visit our ScienceDirect Page: Advanced Engineering Informatics

Keywords:

    Foundation Models
    Large Language Models (LLMs)
    Large Vision Models (LVMs)

Special Issue on AI Trustworthiness and Applications of Large Language Models in Aviation Safety, Accident Investigation and Aircraft Engineering
提出日: 2026-03-31

Contemporary Artificial Intelligence (AI), such as Large Language Models (LLMs) and their variants—Large Vision-Language Models (LVLMs) and Large Audio Language Models (LALMs)—are increasingly adept at performing human-like analysis of large datasets with high efficiency. With proper fine-tuning, AI is expected to revolutionise flight operations, air traffic management, and even accident investigation, thereby enhancing aviation safety and supporting air accident investigations. We can observe that such revolution is going to make great impact on engineering, operations maintenance of aircraft as complex engineered systems, including fault detection and diagnosis in aircraft systems, LLMs for predictive maintenance of aircraft components, explainable AI for avionics and flight control systems, future aircraft cockpit design, and AI in aircraft design optimisation and simulation. However, a critical challenge remains: their trustworthiness.

The AI Roadmap 2.0 of EASA (2023) adopted seven requirements for AI trustworthiness, including: (1) human agency and oversight, (2) privacy and data governance, (3) diversity, non-discrimination, and fairness, (4) societal and environmental well-being, (5) accountability, (6) technical robustness and safety, and (7) transparency. Air operations certificate holders, air navigation service providers and airport operators generally adopt data protection and privacy governance frameworks, which may raise significant concerns regarding data privacy and the handling of sensitive and personally identifiable information.

This problem may be addressed by adopting federated learning frameworks and small learning models that limit training to the knowledge module and feature space, without the need for original data resources. Therefore, this special issue focuses on unleashing the potential of safe and reliable AI in aviation safety, human factors, and accident investigation, while building trust among operators, end-users, accident investigators to achieve safer operations and more efficient accident investigations.

Topics of interest include, but are not limited to:

    Novel methodologies and applications of AI in aviation safety, human factors, accident investigation, aircraft system diagnostics and health monitoring
    LLMs/LVLMs/LALMs and fine-tuning strategies for aviation safety applications, predictive maintenance and fault detection in aircraft components
    Applications of explainable AI in flight operations, air traffic management, accident investigation, avionics, flight control and aircraft design optimization
    LLMs/LVLMs/LALMs-enabled human factors analysis with neuroergonomics for aviation safety
    Trust assessment and quantification on AI applications in aviation
    Theoretical foundations and concepts of human-AI teaming in aviation
    Human-AI interaction design and evaluation for aviation safety, cockpit design, engineering decision support in aircraft operations and maintenance

Guest editors:

Prof. Dr. Kam K.H. Ng
The Hong Kong Polytechnic University, Hong Kong, Hong Kong
(Air Traffic Management, Aviation Safety, Neuroergonomics, Single-pilot Operations)

Dr. Wen-Chin Li
Cranfield University, Cranfield, United Kingdom
(Human-Centric Design on Virtual Flight Data Recorder, Single Pilot Operation Flight Deck Design, Human-System Integration on Digital Tower Operations, Application of Biofeedback/Eye Tracking for Digital Aviation, Cross-Cultural Safety Management and Safety Resilience)

Dr. Chia-Fen Chi
National Taiwan University of Science and Technology, Taipei, Taiwan
(Accident Analysis & Prevention, Taxonomy of Root Cause Analysis, Human Machine Interface Visual Performance)

Dr. Jens Schiefele
Technical University Darmstadt, Hessen, Germany
(Human Machine Cooperation, Aviation, Explainable AI)

Manuscript submission information:

    Open for Submission: from 22 July 2025 to 31 Mar 2026
    Submission Site: Editorial Manager®
    Article Type Name: "VSI: ADVEI_AI and LLMs in Aviation Safety" - please select this item when you submit manuscripts online

All manuscripts will be peer-reviewed. Submissions will be evaluated based on originality, significance, technical quality, and clarity. Once accepted, articles will be posted online immediately and published in a journal regular issue within weeks. Articles will also be simultaneously collected in the online special issue.

For any inquiries about the appropriateness of contribution topics, welcome to contact Leading Guest Editor (Kam K.H. Ng).
Guide for Authors will be helpful for your future contributions, read more: Guide for authors - Advanced Engineering Informatics
For more information about our Journal, please visit our ScienceDirect Page: Advanced Engineering Informatics

Keywords:

    artificial intelligence (AI); 
    large language models (LLM); 
    trustworthiness; 
    aviation safety; 
    human factors; 
    human-AI teaming