A Digital Plant Strategy for Increasing Value using a Real Time Data Infrastructure
One Process Unit Template is used to digitize a process plant.
The business objective is to track the Production Variance or every process unit while tracking the consumption of energy, water and ther consumables.
One Process Event emplate is used to aggregate the production and consumable variable for nterprise tracking for every plant and process unit.
A Master Algorithm or Gross Operational Mode detection generates the Mode of Operation “Running OK” meaning on target, “Trouble” Meaning off target, “Idle” Meaning
No Material Available, and other such as “Maintenance”.
The Plant operational data can be visualized in Power BI as hown in this example or using PI Coresight.
The Algorithm is called Follow the Money by some customers.
One Process Unit Template with Analytics and Event Frame Generation is used to put the data into one Process Unit Event Template. All units are filter by the same algorithm providing a simple way to understand the data and for detecting insights. The data sets are reused for machine learning or for online predictive calculations.
PowerBI Desktop is published to the Cloud to share the information by the functional teams at the plant or enterprise. They can used their phones, tablets or traditional laptop depending of the needs. The generate real time events can generate Notifications to accelerate the resolution of abnormal situations and active resolution workflows by the teams. Details are presented in Chapter 3 and Chapter 4 of the A Journey towards a Digital Transformation Manuscript (work in progress, see PI Square).
This figure shows the published PowerBI in the Cloud enables the use of Cortana to share the information.
Example of Event Frame results Generated Dataset for analysis, visualization and collaboration by the different function in a typical production environment. The following graphic shows the Business Process Workflows in a typical industrial environment. (Kennedy, Bascur, Hydrocarbon Processing). The event frame data set aggregate the real time data into information based on the operational modes (Running, Trouble, Idle, Down or Maintenance).
Business Workflows in a typical plant. This example features the Planning versus Operation and Maintenance workflow. From Kennedy and Bascur published in Hydrocarbon Processing.
One of derived value of using a Generalized Data Hierarchy as shown in this figure is that a data set can be made available to build predictive analytics models. Simple models can be used in PI Analytics to generate softsensors by correlating quality or yield data versus key process variables. (Bascur, A control framework with distributed intelligence, ISA-88-1556)
Basic business strategy to transform data and event into desired information, Data + Events = Information. The IDEA is Infrastructure, Data, Engineering and Analysis (Kennedy & Bascur, AICHE www.nt.ntnu.no/users/skoge/prost/proceedings/cpc6-jan2002/cpc6-all.pdf ).
The Asset Framework becomes your new knowledge coordinator. (Bascur & Kennedy, Measuring, Maximizing and Managing Refinery Performance, Hydrocarbon Processing, page 111).
The IDEA is to track the variance in Real Time and not wait for the end of day or week to make important decisions in the digital age.