Unified Namespace (UNS) Architecture for High-Throughput Factory Data Ingestion and Analytics

Le résumé fourni par la source

Abstract The pharmaceutical manufacturing industry faces significant challenges in integrating diverse data sources due to multi-generational equipment, varied technology stacks, and fragmented communication protocols. These challenges hinder scalability, increase technical effort, and limit the adoption of advanced analytics and AI-driven solutions. This paper introduces the Factory Integration Layer (FIL), a scalable, event-driven platform designed to bridge the gap between Information Technology (IT) and Operational Technology (OT) systems across globally distributed manufacturing sites. FIL leverages a hybrid edge-cloud architecture and a Unified Namespace (UNS) for standardized data organization, enabling seamless data integration and AI-driven innovation. The FIL platform is based on an event-driven, publish-subscribe architecture supported by messaging brokers such as MQTT and Kafka, ensuring asynchronous communication and decoupling between data producers and consumers. It incorporates centralized governance, semantic interoperability, and edge autonomy to address challenges such as equipment heterogeneity, semantic fragmentation, and cybersecurity risks. By implementing real-time data contextualization at the edge, persistent event storage in the cloud, and a centralized control plane, FIL delivers validated failover, disaster recovery, and horizontal scalability. The platform is designed to transform factory data into a shared enterprise asset, enabling a data-product marketplace paradigm with fine-grained access control and compliance. FIL provides the governed, contextualized, and low-latency data infrastructure required to enable enterprise analytics and AI applications across a global manufacturing footprint. This approach redefines factory data integration, turning it into a repeatable enabler of continuous improvement while addressing critical security and governance requirements.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.