Process Industry Professionals: What is your tolerance for bad data & poor information?
Whether it is called “Industry 4.0”, “Edge Computing”, “Big Data” or “IOT”, the tolerance of People, Applications and Business Processes to poor quality data in industrial plants is diminishing rapidly. The Modern process industries thrive on readily usable information and credible data to deliver sustainable tangible business results.
Until recently, Data Quality was relegated to Data Validation and Data Reconciliation projects supporting certain specific functions such as production and yield accounting, mostly in hydrocarbon process industries. But, times have changed — it is time to embrace Data Quality by implementing a strategy that will result in a trustable industrial plant dataset. The targeted outcomes are clear:
- There are three classes of data users we must satisfy:
- Software Applications (downstream from sensors)
- Operational and Business Processes
- Industrial plants are continuously generating different types of data (for instance, streams, events and transactions.) which must be assembled into a coherent, accurate, credible, functional and usable dataset suitable for all class of data users. We call it the “Industrial Plant Dataset”
- To connect the data users with the “Industrial Plant Dataset” in a reliable, sustainable way in order to move operations and business forward and redirect when necessary.
The strategy of Pimsoft is based on implementing Sigmafine, a robust system of checks and balances. Using conservation principles, statistics, engineering standards and calculations to monitor and assemble Industrial Plant data, Sigmafine generates a dataset that is coherent, trustable and usable — a dataset ready for business.
We implement Sigmafine as an adaptive Enterprise Information Service (EIS) which can evolve with the business and operational needs of data users. Ensuring Data Quality is no longer a task or a function relegated to desk of a statistician, a production accountant or to a point solution. Instead, it is an enterprise service that runs autonomously, that spans the value chain of the industrial plant and which is accessible to all data users.