Superior Controls’ Senior Project Engineer, Matthew Martin, discusses the steps a major pharmaceutical company is taking to implement the OSI PI Software Suite.
The generation, maintenance, and use of historical process data is always an important part of the discussion in the pharmaceutical manufacturing sector. It affects things such as regulatory compliance, process optimization, and overall business strategy.
To enhance this aspect of their business, one of our customers has chosen to leverage the OSI PI Software Suite. The old process data is being migrated to the OSI PI system from an older historian software, and all process data moving forward will be collected using the OSI PI system. The OSI PI system is one of the leaders in the industry and provides various architecture and licensing options to accommodate customers’ needs. The functionality of the OSI PI system can be broken down into the following subsections: collecting data, storing data, contextualizing data, and visualizing/accessing data.
Collecting data is the process of taking data from a source and sending it along the pipeline to be stored. In this case, the data sources are PLCs and relational databases. The PI System uses PI Interfaces and Connectors to perform the data collection. OSI offers over 450 interfaces allowing for collection from a vast array of data sources.
In the past the customer had issues with its legacy data collection; data gaps and outages would occur too frequently. PI Interfaces facilitate data collection and allow for redundancy to ensure that there is not a single point of failure in the system. In addition to the available redundancy, PI interfaces also provide buffering: if the archive is unavailable, the interface will store data locally until the archive is available again. The buffered data is then forwarded to the archive, resulting in no overall gap in data.
Storing data is the process of archiving the data that is received from the PI interfaces and connectors. The PI System stores the data in one or more PI Data Archive servers. PI Data Archive servers can be implemented in a collective to further improve availability of the data and the reliability of the system.
Each value archived by the system is defined as a point; to help create these points, OSI provides a Microsoft Excel add-in that allows configuration of points in a very user-friendly and efficient manner. When creating points, the user defines all of the tags’ attributes, identifying the nature of the data as well as how it should be archived.
Settings for exception and compression allow the system to store data in an efficient manner, while also providing the user the flexibility to specify exactly how data points should be stored. The exception criteria affect what values the interface will send to the archive. The PI interface will factor in how long it has been since the last value was sent, and compare the current value to the previously stored value to see if it exceeds the exception deviation setting. Once a value has passed one of the exception tests, it is then sent to archive where it undergoes a compression test. The compression test uses a ‘swinging door’ algorithm to determine whether to archive or discard the value from the interface. When combined, the exception and compression settings allow critical data to be fully retained, while significantly reducing storage requirements.
Organizing and Contextualizing Plant Data
After the data has been securely collected within a robust archive architecture, the next thing to consider is how the data will be used. A unique aspect to the PI System is its ability to allow the user to easily contextualize the data through the use of the Asset Framework feature. Asset Framework allows users to create an organized model of their facility or process area in a hierarchal manner. Creating objects within this hierarchy and allows for parent-child relationships between objects. Another term for this in the industry is creating a digital twin. This is a powerful feature, as it allows data to be presented with more context and relevant information. Asset Framework also supports sending notifications based on user-configured settings, and supports generating event frames to further analyze and structure your incoming data.
Visualization and Reporting
In addition to being able to provide a structure in which data can be contextualized, OSI also offers a robust set of tools for visualizing and making the data actionable to end users:
- For engineers or end users who want to get their hands directly on the data and do their own data analysis work, OSI offers PI Datalink. PI Datalink is a Microsoft Excel add-in that allows users to pull data from the archive or access Asset Framework to bring data directly into Excel.
- PI Vision allows users to create their own ad hoc displays to visualize the data they require to perform their day-to-day tasks. These screens can access data from the PI Archive directly, or can access the Asset Framework structure to further improve the workflow and access to data. PI Vision allows a user to create anything from a high level dashboard of the process for management, down to detailed process graphics for specific equipment sets. It also allows users to create displays leveraging asset swapping for a reusable display for assets of the same type within the AF model.
- RT Reports is a tool that allows for more regulated and formal reporting to be created using the same data sets. In addition to enabling the creation of detailed reports, it also allows enforcing strict version and issuance controls which lends itself well to pharmaceutical use.
These are just a few reasons why this customer has chosen the PI System as their software suite for historical data collection. At the completion of this effort, the PI System will be leveraging 20 interface pairs, one for each process area, to collect approximately 1,000,000 tags. Superior Controls is currently implementing multiple, similar OSI Soft based projects to meet these needs at manufacturing facilities throughout the country.