The use of smart sensors and actuators to improve manufacturing and industrial processes is known as the industrial internet of things (IIoT). Predictive maintenance is one of the most widely advertised advantages of IIoT devices in the industrial business. Organizations can estimate when a machine will need to be maintained using real-time data generated by IIoT systems. Another advantage is that field service is more efficient. Field service technicians can use IIoT technologies to discover possible flaws in client equipment before they become serious problems, allowing them to remedy the issues before they cause customers any difficulty.

However, there are various challenges in deploying the IIoT. For example, adequate infrastructure placement for timely and accurate collection of data from multiple sensors is one of the biggest challenges for IIoT. Other major challenges for IIoT include the aggregation of data at gateways/cloud and the development and adoption of advanced artificial intelligence (AI) algorithms. Hosting machine learning-related tasks at the edge for the purpose of reducing the network load is also difficult in cases of IIoT. Some graph-based machine learning (ML) methods/architectures have been presented in light of the benefits of graph-based ML for IIoT. Despite the fact that these approaches have had some success, they still face a number of scientific engineering obstacles, such as isolated data silos, data inaccessibility, inefficient workflows, poor data quality, and privacy protection.

This Special Issue aims to attract high-quality original research and review articles from academia and industry that address unresolved topics in graph-based intelligence-driven IIoT. Academic and industrial researchers and practitioners are invited to submit high-quality unique work in this area that uses graph-based machine learning/deep learning, data gathering and analysis, online and unsupervised algorithms, robots, cloud computing, etc. Authors are also urged to show how their suggested graph-based intelligence for IIoT solutions can be used to supplement large-scale industrial systems and small-scale laboratory testbeds in the actual world.

Potential topics include but are not limited to the following:

• Applications of graph-based AI for industrial IoT
• Graph-based machine learning for predictive analytics including maintenance
• System aspects of the data collection pipeline
• Graph-based AI-driven private networks for IIoT
• Large-scale graph-based AI integration in industrial environments
• Applications of graph-based intelligence for control functions and operations
• Applications of computer vision in IIoT
• Graph-based unsupervised, self-supervised learning and online/offline reinforcement learning for IIoT
• Graph-based modeling and learning and representation learning for IIoT
• Transfer learning for IIoT including real-world implementations
• Testbed and performance evaluation studies of graph-based AI/ML techniques for IIoT
• Computing architectures supporting graph-based AI/ML techniques for IIoT
• Graph-based AI-/ML-driven networking and wireless communications for IIoT
• Sustainability and carbon neutrality with enlightened IIoT