Graph-Guided Neural Networks: Revolutionizing Anomaly Detection in Chemical Engineering

The increasing complexity of modern chemical engineering processes presents significant challenges for timely and accurate anomaly detection. Traditional model-based and statistical methods often struggle to capture intricate variable interdependencies and nonlinear dynamics in large-scale processes. With the emergence of advanced machine learning, graph-guided neural networks have gained attention for their unique ability to represent structured data and relationships among process variables. By integrating graph theory with deep learning, graph-guided neural networks offer a promising framework to model spatial-temporal dependencies, enhance interpretability, and detect subtle or early-stage anomalies that conventional methods might overlook. As chemical processes become increasingly digitized and data-rich, the need for intelligent, structure-aware anomaly detection methods has never been more critical.

This Research Topic aims to bring together cutting-edge research and practical advancements in the development and application of graph-guided neural networks for anomaly detection in chemical engineering. Our goal is to showcase how graph-based models, including graph convolutional networks, temporal graph networks, and attention-based frameworks, can transform the way anomalies are identified, interpreted, and addressed in both batch and continuous systems.
We seek contributions that span from theoretical innovations and algorithmic design to real-world applications in areas such as fault diagnosis, safety monitoring, quality control, and predictive maintenance. Emphasis will be placed on methods that go beyond black-box modeling by incorporating domain knowledge, ensuring interpretability, and demonstrating generalizability across different processes. We also welcome interdisciplinary works that connect chemical engineering with data science, control theory, and AI.
Through this collection, we aim to foster a collaborative platform where researchers and practitioners can explore the full potential of graph-guided neural networks in creating safer, smarter, and more resilient chemical engineering systems.

This Research Topic invites submissions on both fundamental and applied aspects of graph-guided neural networks for anomaly detection in chemical engineering. Topics of interest include, but are not limited to:
• Graph construction and representation learning for process data
• Graph neural network architectures for anomaly detection
• Integration of graph-guided neural networks with control systems or digital twins
• Case studies in fault diagnosis, monitoring, or predictive maintenance
• Benchmarking graph-guided neural networks against traditional or other AI-based approaches
• Interpretability, explainability, and domain knowledge integration

We welcome Original Research articles, Comprehensive Reviews, and Perspectives that contribute novel insights or summarize the state of the art. Submissions should demonstrate clear relevance to chemical engineering applications and highlight the advantages of graph-based modeling in anomaly detection.

Article types and fees

This Research Topic accepts the following article types, unless otherwise specified in the Research Topic description:

• Brief Research Report
• Editorial
• FAIR² Data
• FAIR² DATA Direct Submission
• Methods
• Mini Review
• Original Research
• Perspective
• Review
• ... View all formats

Articles that are accepted for publication by our external editors following rigorous peer review incur a publishing fee charged to Authors, institutions, or funders.

Article types

This Research Topic accepts the following article types, unless otherwise specified in the Research Topic description:

• Brief Research Report
• Editorial
• FAIR² Data
• FAIR² DATA Direct Submission
• Methods
• Mini Review
• Original Research
• Perspective
• Review
• Technology and Code

Important note: All contributions to this Research Topic must be within the scope of the section and journal to which they are submitted, as defined in their mission statements. Frontiers reserves the right to guide an out-of-scope manuscript to a more suitable section or journal at any stage of peer review.