Real Time Operational Dashboard Case Study

The Dashboard Spy is pleased to be cited as a enterprise digital dashboard resource in the fine whitepaper by Sidney Shek titled Enterprise Digital Dashboards as Rich Internet Applications. This CSC paper was added to the CSC Leading Edge Forum Knowledge Library a while back, but came back to mind recently as I helped a Dashboard Spy reader design an operational dashboard.

Let’s take a look at a prototype dashboard featured in the whitepaper. It’s a real-time operational dashboard designed to track the operation of a Hot Strip Mill. You may have seen videos on Mr. Rogers where slabs of steel are heated and rolled through a series of operations until you have an inch thick piece of metal. It’s an awesome process that deals with huge machines and temperatures of up to 2300 degrees farenheit.

hot strip mill steel slabs Steel Mill Process Rolled Steel Mill

Here is the Hot Strip Mill dashboard itself.

Hot Strip Mill Dashboard

Let’s have a look at the commentary on this dashboard from the white paper. Here is the use case:

For this example, this dashboard is designed to be used by a manager of a production plant to view what is currently happening in the plant. The dashboard can be kept running on his/her computer on a second monitor in the background so that the manager can glance at it occasionally and pick up key performance indicators and current production status. When problems in the plant occur, there is sufficient detail on the dashboard to identify possible causes (through operator comments) and possible impacts (through information about the most recent products). The manager can drill down from the dashboard to investigate further if required. Changes in the plant’s status (e.g. when a stoppage occurs) are pushed from the server through to the client dashboard so that the information on the dashboard are kept up-to-date without requiring frequent polling of the server.

Click on the “read more” link to see the rest of the commentary on this real-time operational dashboard.

Here is a tour of the dashboard:

Current production status information is present in the top left corner, as this is considered the most prominent part of a display.

The plant’s current performance is shown by KPIs for the current shift (‘Current Shift’panel). The tables show the current values (and projected values for the shift, if applicable), and a linear gauge that compares the current (or projected) value to the plant’s KPI target ranges. See below for more detailed discussion of linear gauges.

The ‘Recent Events’ panel shows recent operator comments and delay information to provide context to the plant’s current operation. Managers can drill down into the comments for more detailed analysis by double-clicking the appropriate row.

The ‘Latest Produced Coils’ panel shows key information about recently completed products, including whether the product met its specifications (and hence can be sold).

Products failing to meet specifications have a red cross in the ‘In Spec?’ column to highlight the problem to the manager. The ‘Recent Events’ information may indicate 42 possible causes of out of specification products which may trigger further investigation.

For more details on specific products, managers can double click on the product to drill down into a detailed report (provided by an existing reporting system). The URL is currently generated in code on the client side based on data obtained from the backend (e.g. the product identifier)

The ‘RSS News’ reader panel has industry related news, which provides context around how the business is operating compared to the rest of the world. Users can change the current RSS feed by clicking on the ‘new feed’ icon, and selecting a feed from a predefined list, or users can enter a URL for the feed. This is an example of providing useful configurability to end-users.

The ‘Safety News / Plant Manager’s Comments’ panel is an existing web application showing a message board for all operators and users. This is a basic example of integrating a Flex application with HTML/Javascript web applications using a custom Iframe Flex component. The I-frame is capable of resizing and moving as the Flex application moves, resizes and scrolls.

There is space for ‘Live’ video feeds in the bottom right corner of the dashboard. There is an increasing ability to bring live video to the desktop with the latest generation of network IP video cameras, MPEG-4 and more powerful client PCs. This would allow managers to actually see how the plant is operating in addition to the numeric or textual information. A possibly useful feature would be to allow users to switch the video feed to full screen mode, or pop up in a separate window.

The layout is such that the most important information on the dashboard can be seen even on a standard 1024×768 resolution display to accommodate most users. All information can be seen on a slightly larger 1280×1024 display.

Users can print a copy of the dashboard by right-clicking and selecting the ‘Print Dashboard…” option. This will bring up a standard Windows printing dialog box. The dashboard is scaled to fit onto the page size and orientation that the user selects.

The use of the linear gauge for the data visualization is explained:

Visualisation with linear gauge

The linear gauge is based on the ‘bullet chart’ described in section 6.1, and is modelled similar to an audio level meter. The vertical bar will slide along the horizontal axis showing the current 43 value by its location according to the scale on the horizontal axis. The bar changes colour based on which KPI target range that the value is in (red for poor, yellow/orange for okay and blue for good). The target ranges are shaded in the background of the chart, with the target value shown by a dark blue target circle.

Linear versus radial gauges

It would be preferable to use a radial gauge to display the KPI value in comparison to target ranges as users can quickly and easily interpret the displayed values [6.1]. However, for this example business case, eight KPIs needed to be shown and it was too congested with eight radial gauges. In contrast, eight linear gauges could fit nicely into the space, and hence these were used.

For more, please see the paper. It’s the best resource I’ve seen in a while.

The Dashboard Spy

Tags: Hot Strip Mill Dashboard, Real-time Operational Dashboards, business intelligence data visualization

Have you visited this Dashboard Spy site? Dashboards

Leave a Reply