Explore volumes of production data with easy, search-engine-like abilities to monitor and improve operations in ways you haven’t even thought of yet.
By Andrew Ellis, manager, Information Software Technical Consultants
Every second, nearly 40,000 Google searches help us find information online. It’s become so innate that we simply say we’re going to “Google it” when looking for answers; “Google” has become a verb. Yet this power stops at the front door of a production facility. Industrial firms haven’t had a similarly robust search capability for their operational data.
As these firms connect more of the things and systems in their plant and rely more on the information they produce, search functions seem like a natural fit. But if you’re like most companies, you still depend on pre-programmed dashboards and reports to access this information.
Fortunately, search technology is advancing into the industrial enterprise. All the connected technologies spinning off data are available to analytics software, and their value could be improved through search with natural-language processing. In other words, workers can scour their industrial operations like they do the Internet.
Easily searchable data provides insight into operational performance and helps workers quickly diagnose and resolve production issues.
Beyond the Dashboard
The current state of most industrial software involves dashboards and reports configured to deliver specific pieces of information to specific workers. Getting additional details out of the defined data sets often entails additional programming and design. That’s a time-consuming task that requires specific skill sets.
Configuration time leaves employees in the dark, when what they need is the information to resolve an unexpected issue. It can also shortchange workers of potentially valuable insights that simply weren’t built into a report or dashboard.
Searchable analytics can solve these problems. Once data sets are defined, users can quickly search for them — even if they haven’t been previously visualized. This transforms users into self-serving data scientists, giving them the ability to see a broad view of their operations through dashboards, and then use search to look beyond those dashboards for additional insights.
A production manager, for example, can track his plant’s overall energy usage on a dashboard. If he sees a spike in consumption, he can then search for energy usage by machine, batch or shift to identify the culprit.
How It Works
Search begins with identified data sets, which are established as part of any Industrial Internet of Things (IIoT) system setup. The data sets include two key elements: what you’re measuring and how you’re measuring it.
Once your data sets are established, users can hypothetically access them. They just need to be able to identify what’s relevant.
This is where natural-language processing is key. It doesn’t require that workers know structured query language, which is what’s used when programming a system to pull specific information. Instead, personnel can simply search using natural language, like we all do in a Google search.
Whether workers want to dive deeper into existing analytics or investigate an issue, they can search simple questions such as: How much energy was used by X machine this month? What was the output by shift this week? What was the top source of downtime in the last 24 hours?
Using the Search
Searching for data is one thing. How you use it is another.
Certainly, there’s inherent value in being able to scour the wealth of your production data. It can give you ad-hoc insights without the hassle of creating new dashboards or reports. It can help you more quickly diagnose and resolve issues. And it can help you instantly answer questions that arise in production meetings or from shift reports.
However, advanced analytics let you do far more with your search results. For example, you can add search results to reports or dashboards with a single click. This allows you to extend your existing analytics with valuable new insights as you discover them. Users can also copy existing dashboards, make changes and create their own display.
A Siri for Your Plant?
Searchable analytics for industrial purposes are still in their infancy, but they’re already giving way to other advances we enjoy in our daily lives.
Some manufacturers are now testing analytics with voice recognition. Just like many of us get information from Siri®, Alexa and Google by verbally requesting it, production workers might soon be able to do the same thing on the plant floor.
This has the potential to provide plant-floor workers with even faster and more convenient operational information. A maintenance worker fixing a machine, for instance, could ask for diagnostic data without taking off his gloves or putting down his tools. And workers in controlled environments wouldn’t need to touch a screen or keyboard to confirm recipe details.
Voice recognition also could help industrial producers ease increasing workforce woes. Tech-friendly millennials are the leading users of voice-enabled digital assistants. Bringing such technology to the plant floor is just one way to show these sought-after workers the new digital face of manufacturing.
Dive into Your Data
Dashboards and reports give you a valuable but limited glimpse of the data being produced by your manufacturing operations. With new search capabilities, you can look beyond your prebuilt analytics and tap into the vast volumes of used and unused data. You can explore this data to monitor and improve operations in ways that you likely haven’t even conceived of yet.
The Journal From Rockwell Automation and Our PartnerNetwork™ is published by Putman Media, Inc.