The initiative addresses the difficulty of maintaining data control and model integrity in highly constrained field equipment. By leveraging Red Hat’s open source building blocks, Sopra Steria creates an integrated, accredited system designed to operate autonomously. This approach allows critical infrastructure to maintain functionality even when disconnected from central servers, ensuring that decision-making remains local and secure.

The framework utilizes Red Hat OpenShift AI for centralized model training, while Red Hat Device Edge and Edge Manager handle the deployment and maintenance of models on hardware with limited power. This creates an end-to-end automated pipeline capable of managing large fleets of devices. Practical applications include real-time anomaly detection for transport logistics, diagnostic filtering for public health terminals, and distributed inference for regulated signal processing. By standardizing the AI lifecycle, the companies aim to bridge the gap between initial development and the practical realities of industrial field use.