A New Era of AI Integration in Connected and Intelligent Systems

This white paper sets out a vision for a new era of AI integration in Connected and Intelligent Systems (CIS), proposing an infrastructure AI software layer paired with a trust-based, distributed governance model to enable industrial AI to be embedded in existing systems and operational processes. Targeting domains such as surgical robotics, autonomous driving, UAVs and counter-drone systems, defence, infrastructure inspection and emergency response, it argues that pre-trained industrial models, together with small-scale and agentic AI, require a trusted and efficient cloud-edge infrastructure built on a common data model, cloud-native edge paradigms and API-first architectures. The paper describes a shift from traditional embedded engineering towards software-defined, app-based systems running on a metaOS that abstracts the underlying hardware, opening the way to an open app ecosystem and a new app economy, illustrated through the case of smart bidirectional charging across the energy, mobility and buildings domains. It highlights the growing role of AI-enabled edge computing and contends that, without a harmonised reference architecture and multi-party governance model (depicted as an “hourglass”), industrial AI will not scale. Finally, it outlines the CSA CEI-Sphere’s intention to introduce a trusted labelling scheme for dynamic software components such as AI models and firmware updates, and calls for a broad ecosystem uniting system engineers, AI developers, industry and research.