Expanding Data Accessibility from PI Systems
Over the past ten years, businesses have invested considerable effort and resources into integrating data into PI systems. They have leveraged the superior contextualization capabilities of these systems for equipment (Element and Attribute), time (Event Frame), and material movement (Transfer). This has resulted in significant advancements in paperless operations and dashboard visualization.
However, the desire to do more is growing!
Extracting data from PI and other manufacturing systems to integrate with other data sources, a process commonly referred to as IT/OT convergence, has posed significant challenges for many organizations. Third-party service providers often concentrate on extracting a subset of data from the PI ecosystem, but in doing so, they frequently lose important context. Moreover, it has been a complex task to supply both edge nodes and cloud-based systems with high-quality data. Concerns about scalability and reliability also arise, particularly when transitioning from a robust on-premise installation with high availability and automatic recovery features.
So how do you make your data expansion successful?
1. Steer clear of daisy chaining your data pipeline: Every component in sequence can and will fail, reducing the overall system reliability. Clustering and high availability are key.
2. Don’t forget your meta data: Raw sensor data need context, event frames and transfers are key for equipment, process and material modeling.
3. Avoid technology lock in. If your connector only supports a limited number of destinations, you will at some point face roadblocks and work arounds.
4. Make your PI\AF source system cloud ready: If every plant has different data models and naming conventions, it will be difficult to leverage common data structures.
5. Architecture and infrastructure are just the means to the end: Start with selected use case and make them successful. Start with edge or on premise solution and expand to the cloud.
Let us know what you think at sean@processdataexperts.com
