Editorial for the special collection on computational intelligence for industrial automation and optimization

In an era defined by rapid technological transformation, industries across the board—from electric power and cement to petrochemical engineering and green energy—are undergoing a paradigm shift. The digital revolution, once associated primarily with consumer-facing applications, is now deeply embedded in the core of industrial systems. Enabled by advancements in sensing, computing, and networking technologies, industrial sectors are becoming increasingly reliant on artificial intelligence (AI) and machine learning (ML) to automate processes, enhance operational efficiency, and derive actionable insights from vast and complex datasets.

This special collection reflects the emerging convergence of AI and industry, focusing on intelligent automation and optimization from the device level to the enterprise value chain. The data-rich environments within modern industrial operations, encompassing everything from process planning and machinery control to trading workflows and after-sales services, present both unprecedented opportunities and significant challenges. Harnessing this data effectively requires novel theoretical frameworks, robust algorithms, and scalable architectures that can navigate the intricacies of industrial systems while delivering tangible performance improvements.

From the 18 papers submitted to this special collection, 6 were selected following rigorous peer review. These contributions represent cutting-edge research at the intersection of AI, automation, and industrial engineering. Collectively, they offer a multifaceted perspective on how AI-driven solutions can be tailored to address real-world challenges in dynamic, high-stakes environments.

Lurker: Backdoor attack-based explainable rumor detection in online media (Lin et al.) introduces a novel backdoor-based, explainable graph neural network model for rumor detection in online media. By identifying and exploiting propagation structures, this work bridges the gap between cyber-physical systems and information trustworthiness—critical for maintaining industrial and public safety.

Attention-based handwritten Chinese recognition (Xiao et al.) enhances document digitization for power grid maintenance. The integration of spatial attention mechanisms and LSTM architectures achieves high recognition accuracy across diverse handwriting styles, supporting efficient digital record management in large-scale infrastructure systems.

Multi-task transformer-based process monitoring (Wang et al.) proposes a multi-gate mixture-of-experts model for predictive process monitoring, adept at capturing complex event sequences in manufacturing. This contribution exemplifies how multitask AI architectures can improve foresight and adaptability in process-intensive domains.

A backtracking heuristic for strip packing (Li et al.) delivers a practical and efficient algorithm for two-dimensional strip packing—a classic optimization problem with direct implications in manufacturing logistics and resource utilization. The simplicity and performance of this heuristic make it particularly valuable for large-scale, real-time applications.

Neuro-fuzzy and optimal PI control for DC–DC converters (Süleyman and Mumcu) presents a hybrid control scheme for solar power systems, integrating human expertise and optimization algorithms. This work advances energy efficiency and stability in renewable energy systems—a cornerstone of sustainable industrial transformation.

CPBA-CLIM for hazardous chemical incident management (Du et al.) introduces an advanced entity-relation extraction model for building knowledge graphs from incident reports. By combining content-position attention and cross-label matching, the model supports structured decision-making in high-risk, information-dense environments.

This Special Issue serves as a timely and important contribution to the evolving field of industrial intelligence. At its core, it explores how data-driven approaches, particularly those powered by artificial intelligence and machine learning, are revolutionizing traditional industries. The featured studies span a wide range of technical domains—ranging from predictive process monitoring and intelligent document recognition to advanced control systems and knowledge graph construction—all unified by a common goal: to build more adaptive, autonomous, and intelligent industrial systems.

By bringing together research that spans process-level optimization, system-level coordination, and enterprise-level decision-making, this issue highlights the holistic potential of AI to reshape industrial operations from the ground up. The selected papers not only advance state-of-the-art algorithms and architectures but also demonstrate real-world relevance by addressing pressing challenges such as production volatility, energy efficiency, safety, and predictive monitoring.

We hope this special issue will not only inform but also inspire further exploration into intelligent automation and optimization. By offering both theoretical insights and practical frameworks, we believe this collection will serve as a valuable reference for the scientific community, a knowledge base for practitioners, and a roadmap for future innovations in smart manufacturing and industrial digitalization.

We extend our sincere gratitude to all authors, reviewers, and contributors whose dedication and scholarly efforts have made this special issue possible.