The proliferation of Cyber-Physical Systems (CPSs), including the Internet of Things (IoT), is changing our lives. Next generation CPSs aim to conduct pre-competitive research on architectures and design, modeling, and analysis techniques for cyber-physical systems, with emphasis on practical applications. These applications include transportation systems, automation, security, smart buildings, smart cities, medical systems, energy generation and distribution, water distribution, agriculture, military systems, process control, asset management, and robotics. Recently, the emergence of embedded and ubiquitous cyber-physical applications based on IoT, Artificial Intelligence (AI), and 5G have driven the evolution of CPSs. Due to the progressive transformation from host-centric networking to information-centric networking, CPSs pose fundamental challenges in multiple aspects, such as heterogeneous data generation, efficient data sensing and collection, real-time data processing, and greater request arrival rates.

AI will drive the CPSs technology revolution. However, the big and heterogeneous data are changing the future of the CPSs, which should be processed with advanced data management and analytics models. Recently, various learning architectures and techniques, i.e., machine learning, representation learning, deep learning, and transfer learning, have been introduced to revolutionize big data mining and information processing methods. Both traditional learning methods and advanced learning methods are essential to meet the needs of CPSs data acquisition, storage, management, processing, and analysis. In light of this potential, this special issue provides a venue for promoting next generation CPSs based on diverse learning methods. Considering the benefit of learning methods for CPSs, various learning technologies and frameworks have been proposed. Potential applications include the intelligent environment, smart transportation, intelligent energy management, smart health, and big-data-driven urban planning. Even though these approaches have achieved certain success, there exist various scientific and engineering challenges including software and hardware development, computational complexity, data multi-source heterogeneity, and security/privacy problems. These open issues call for extensive attention from both academia and industry.

This Special Issue aims to solicit high-quality original research papers, which address the cutting-edge theories, models, and applications for next generation industrial CPSs, supported by learning methods. Review articles are also encouraged.

Potential topics include but are not limited to the following:

• Learning methods for CPSs
• Learning methods based security, integrity and privacy solutions for CPSs
• Learning methods based 5G communication for CPSs
• Learning methods based intrusion detection/prevention techniques
• Learning methods based physical layer design techniques for CPSs
• Learning methods based energy-aware industrial management solutions
• Learning methods for smart city/grid/healthcare
• Learning methods based architectures, designs and applications
• Learning methods based recommender systems leveraging CPSs
• Learning methods based advanced data analytics for cloud-integrated CPSs
• Integration of AI into CPSs
• Applying learning methods to industrial scenarios