Best Practices on Process Graphics and Ways to Present and Share Information
"With all the data that we have available, the amount of information to present in graphics can be overwhelming," Peter said. "As such, we need to have a hierarchy and a presentation plan on how to go about it."
One of the first objectives is to make sure the data is accurate for the proposed dashboards and process displays. First, the data will be captured, then the data is validated to ensure that sensors are working and sending data properly. Next comes further validation to ensure that the data is within operating limits.
Later in the EIDI implementation, offline tools can be used to holistically monitor the process. If the data is within operating limits (sometimes called operating envelopes), physical or empirical models can be used to generate estimates based on current measurements. Process models can be used for this purpose. Because the EIDI system can display future-time projections indicating what values these models are predicting, it will help operators see variation in actual versus expected results. This last step will not be deployed during the initial phases of the EIDI initiative.
It always a good practice to define end users within a system. Who will be using this data, and what is his or her job? The vital ingredient of an infrastructure is that the person in each particular role is the most important element. "In essence," explained Peter, "we are building a cockpit with a data dashboard to simplify the job." Employees should be able to draw a picture or utilize an existing picture from a library. They should be able to easily access trend data so that they have an immediate understanding of refinery conditions. However, people can easily be distracted by color and movement. Flashing colors in abundance become distractions and may cause refinery operators or maintenance workers to lose focus.
Three principles apply to real-time data visualization:
- 1. Clarity
- • Graphics are easy to read and intuitive.
- • Graphics show the process state and conditions clearly.
- • Graphics do not contain unnecessary detail and clutter.
- • Graphics convey relevant information, not just data.
- • Information has prominence based on relative importance.
- • Indications of abnormal situations are clear, prominent, and consistently displayed.
- • Graphics make people aware when information is not being updated.
- 2. Consistency. The colors used by the background and the real-time graphics follow standard, consistent norms, such as green reflects normal/desired range values; red values are an alarm or undesirable condition; flashing values need immediate attention. Consideration needs to be given so that color-blind people can still effectively use the displays.
- 3. Feedback. This refers to the active interaction of a touchscreen or mouse and additional access to detail information. An approach to avoid graphic cluttering is to hide detailed information until the user clicks on an object. However, if the user wishes additional information, the human interface will respond with configured feedback as embedded information in the graphical visualization objects. For example, if you touch a specific trend, additional information such as the current value, minimum, maximum, and rate of change for the selected time period is displayed.
For dashboards, a recommendation is to locate the key metrics on the top, then follow with trends or charts representing the data. If required, present a table with summarized data. For process flow diagrams, it is important to show the data starting at the left of the process supply chain and ending at the right with the results.