3 Tips for Building Better Reports
Focus on Parameters, Data Structures and Delivery5 minute read Download PDF
Effective reporting is very important to an industrial organization: it helps to measure operational performance, identify areas for improvement, and increase accountability among team members. But any report is only as valuable as the data that goes into it. That’s why the key to building more effective reports is not to add fancier charts and graphs but to focus on the data and delivery methods you’re using. These basic factors are critical for effective reporting, regardless of which reporting system your organization currently uses.
Tip 1: Use Parameters to Save Time and Customize Reports
For any report, you need to specify one or more data sources, which can include queries, scripts, static CSV and Excel files, and web services. The most effective way to get data from a source is to create a set of parameters.
Parameters define the data sources and set values for a report. All parameters have a name, a parameter type (such as date, string, long, double, Boolean, or dataset), and a default value.
Parameters come in two forms: data parameters and filter parameters. To show values on a report, use a data parameter, which is linked to the current value of a tag, such as “Tag/Time.” To retrieve data, use a filter parameter such as “StartDate” and “EndDate.”
Parameters save a lot of time and effort, and allow you to customize reports, because they make it possible to build a report once and use it over and over. A report with parameters can reference different information even though it is defined in one place. For example, if you’re running four machines and need a data report for each one, you should use parameters so that the report can get data from any of the four machines. If you don’t use parameters, you will have to create four different reports of the exact same data and hardcode the machine that you want information from.
Tip 2: Choose the Right Data Structure
Data structures are used to put together reports. There are three main types of data structures: single dataset, peer datasets, and parent-child dataset.
Single dataset: The most basic data structure is a single dataset comprised of simple data from one source which is displayed in rows and columns. When you only need to display one dataset, this is the data structure to use. For example, use a single dataset when you query a historical source and get one result back. You can create a report of your history with no other data on it, just a simple dataset displayed in a graph or table.
Peer datasets: You should use peer datasets when you have multiple sets of non-relatable data that you need to display together in the report. These data structures are also useful when you don’t know how long the datasets are going to be but you want to show them one after the other so that it will be easier for the reader to take in the data.
As an example, a recycling facility might need reports for the different products it made that day. By using peer datasets, the report could show all the data about the plastic recycling for the day, followed by all of the aluminum recycling for the day, and so forth. This report could contain different columns for the different datasets since the values you get back for aluminum may be different than the values for plastic. The datasets are not related except for the fact that they are all from products made on the same day.
Parent-child datasets: When dealing with related data, it’s best to use parent-child tables. In this data structure, child tables are inserted into each row of the parent table. This is also referred to as nesting because one set of data can be nested inside of another; in other words, the child’s data is nested inside the parent’s data. This structure lets you gather data from different sources in a way that shows the relationships between the data.
For example, suppose your report is set up with a row for each piece of equipment that you need to list. You could add a separate child table under each row that shows some additional information, such as downtime, for that piece of equipment.
Tip 3: Determine Your Report Delivery Method
Once you’ve defined and structured the data for your report, you have to make sure it gets to the people who need it. There are basically three ways to deliver a report: manually, automatically, or through a specified event that triggers the report.
Manual: When you need to generate a report quickly or spontaneously, you can manually fill in the data and parameters to get a report back that you can print on the fly.
Automatic: These reports are scheduled at a set, recurring time, such as hourly, daily, weekly, monthly, etc. Use this delivery model when you know that you’re going to need the same info at the same time on a consistent basis. For example, you may need to get a report every workday at 5:00 PM to find out how your operations performed that day.
Once an automatic report has been scheduled, select whether to deliver it by emailing it, printing it, saving it to a server, uploading it to a website, or through some combination of these methods.
Triggered: This is also known as event-based reporting because this type of report is triggered when a certain event occurs in the PLC, such as a tag change, a shift change, the finish of a production run, etc.
For example, a widget manufacturer is packing widgets into boxes. Whenever a box gets filled up, the PLC recognizes that it has completed filling that box and automatically prints a Workin-Process (WIP) label. The WIP label was generated automatically based on the event of the box getting filled – it did not require a human worker to push a button at a certain time to generate the label. The automatically printed WIP label can be placed on the box and then the box can continue on its way.
As with automatic reports, you have to specify how triggered reports are to be delivered, whether by emailing, printing, saving, uploading, or some combination.