Artificial intelligence for robust and reliable networks in cyber-physical systems

The field of computer science is witnessing a rapid integration of artificial intelligence (AI) within cyber-physical systems (CPS), particularly as these systems underpin complex networks in domains such as smart infrastructure, communication networks, transportation systems, and connected healthcare environments. While AI’s promise is substantial, its deployment in networked CPS presents unique reliability challenges. Issues such as network-induced latencies, intermittent communication links, sensor drift, adversarial environments, and fluctuating resource availability raise critical questions about the continuous dependability of AI-enabled decision-making. The latest research highlights both the progress in distributed, adaptive, and edge AI—but also persistent uncertainties about robustness, resilience to communication disruptions, and integration with dynamic multi-agent or distributed systems. Despite advances in AI-driven analytics for networks, there remains a pressing need for frameworks and methodologies that ensure robust, reliable, and trustworthy AI in the face of real-world network variability. This Research Topic aims to advance the understanding and development of robust and reliable AI technologies specifically within the context of networked cyber-physical systems. The focus is on ensuring consistent AI performance under real-world networking constraints, such as distributed topologies, fluctuating bandwidth, packet loss, synchronization issues, and adversarial network conditions. Through this Research Topic, we seek contributions that critically investigate the intersection of AI reliability with the architectures, protocols, and standards of modern communication networks. Key objectives include identifying theoretical foundations, practical challenges, and novel solutions that address reliability gaps in AI when deployed within operational networked environments. The scope of this Research Topic is limited to themes at the intersection of AI reliability and networked infrastructure within cyber-physical contexts, as distinct from topics centered purely on traditional network management or abstract AI ethics. To gather further insights in this targeted domain, we welcome articles addressing, but not limited to, the following themes: - Robust AI for distributed, edge, and federated CPS networks - AI reliability under variable communication quality and intermittent connectivity - Network-aware AI calibration, self-healing, and resilience strategies - Validation and benchmarking of AI models in networked real-world environments - AI-based sensing, data fusion, and decision-making in interconnected sensor networks - Measurement, monitoring, and runtime assurance of AI performance in networked CPS - Frameworks for trustworthy, explainable, and standards-compliant AI in distributed infrastructures We accept original research articles, comprehensive reviews, methodological studies, and real-world case studies. Topic editors: Irida Shallari, Vincenzo Gallo, Mattias O'Nils Source: https://www.frontiersin.org/research-topics/81180/artificial-intelligence-for-robust-and-reliable-networks-in-cyber-physical-systems