Introduction to the Special Issue on Artificial Intelligence and Cyber-Physical Systems: Part 1

By using a combination of machines, sensors, embedded computational intelligence, and various communication mechanisms, Cyber-Physical Systems (CPSs) monitor and control physical elements with computer-based algorithms, capable of autonomously reacting to and affecting their physical surroundings. Advances in CPS should enable capability, adaptability, scalability, resilience, safety, security, and usability far beyond what is available in the embedded systems of today. In light of the rapid advancements in artificial intelligence (AI) and communications, there is an increasing demand for these intelligent CPSs, such as connected and autonomous vehicles that monitor and communicate with their surroundings and smart appliances that optimize energy consumption based on environment and occupant behavior. To realize the vision of AI-enabled CPS, there are several research areas we can expect to come to the fore. For example, new methods to combine data-driven machine leaning and model-based learning for decision making and real-time control of cyber-physical systems are very promising. Meanwhile, traditional ideas in CPS research are being challenged by new concepts emerging from AI and machine learning. For example, what do high confidence and assurance mean in the context of autonomous systems that learn from their experiences? How does one address the trinity of challenges of trustworthiness, resilience, and interpretability of artificial intelligence in its integration with high-assurance cyber-physical systems? How does one reconcile the concepts of machine learning and data-driven modeling with approaches used in model-based design and formal methods? To explore these new directions and address new challenges, this special issue features 12 articles on the topics of AI and CPS.