For example, for condition monitoring, we used to make rounds and enter data manually. Now we can gather much more information automatically. "Sensors are more accurate than people with clipboards, and they provide real-time data," said Bonz Hart, CEO, Meridium. "Their size and cost are going down; they're now less than 30 cents apiece. But they can generate 63 million readings per year times a thousand—or 10,000—sensors per facility.'
We can store that much data, but how can we evaluate it? And why would we want to? "A teenager generates hundreds of gigabytes of text messages, and we can store those, too, but they're not very useful," Hart says, "Unless the teenager goes missing. Then suddenly they're very useful."
It's pretty easy to see how industries can benefit from leveraging sensor data, but it takes expertise on both the information technology (IT) and operations technology (OT) sides, one more reason for both to be on the same side. "The enabling technologies are developed for commercial applications. Now we're adopting them for industry," says Khris Kammer, information partner and competency manager, Rockwell Automation. "This is accelerating the convergence of IT and OT, allowing people to use data in unprecedented ways."
We also tend to focus on what operations can serve to management, but IT/OT collaboration goes both ways, promising to finally close the loop on plant performance.
"The availability of information from plant assets is driving IT/OT connectivity," says Mike Brown, process application consultant, Honeywell Advanced Solutions. "Today we take ERP information into the plant, often manually. IT/OT integration increasingly lets us bring that information down automatically, but also back up to measure plant performance and support decisions, by telling us what's actually happening. By measuring productivity and yield at the automation layer, comparing plant actuals to ERP projections and making real-time adjustments, we can close the loop on the ERP system."
Along with optimizing performance, we can truly understand how the condition of our assets might prevent us from achieving it. "The great increase in computing power and storage means we can take more information from instrumentation, aggregate it and integrate it to get a full view of risk," says Roy Whitt, senior vice president and Asset Answers General Manager, Meridium. "We can close the loop on maintenance."
But that's just the beginning. "It's easy to imagine how sensors will provide feedback to industrial facilities for analysis and prediction, to predict failures and prevent shutdowns," Whitt says. "But like people couldn't imagine how the Internet would affect their lives, we still can't imagine all the ways this will affect operations and maintenance. The information we'll be getting through the IIoT – from production data to inventories to individual components, parts, bolts and seals – we still can't imagine how that will change the way we run our plants."
With IIoT, "It's not a killer app. It's a combination of enabling technologies – software, communications – with people and a process, to get payback," says Barry Johnson, global director, information solutions and process business, Rockwell Automation. "You may not need the infrastructure without IIoT capabilities, but with them, you do. You need good, solid ways to collect, move and analyze data."
Build the infrastructure, mind-meld IT and OT, and the applications will come. It's revolutionized the commercial side, and the revolution is poised to roll through industry in ways we can't yet imagine.
"In the Internet of Things, the Internet is not the big deal. It's what the things can do. When we stick the IT in the thing, we get exponential improvement," says Michael Porter, professor, Harvard Business School. "Data is becoming one of the factors of production – land, labor, capital, and now, data." We're moving up the continuum from monitoring to control to optimization and now, autonomy.
Leveraging IIoT requires folks who have an inkling of which data matters, how to analyze it and what to do with the results. "To take advantage, we need new skillsets," says Porter. "Along with the existing MEs and EEs, we need data scientists, system engineers, developers and systems integrators."
Those skillsets may appear new to Porter, but not to you.
