Optimizing Control for the Industrial Internet of Things

Whether you see the Industrial Internet of Things (IIoT) as exciting or disrupting or all just a bunch of hoo-ha, you would be better served to get down out of the clouds (so to speak) and really look at what’s needed to start the journey. It’s not as productive to talk about the technologies involved as it is to talk about outcomes, and it’s time to give customers what they need to start becoming more productive and efficient.

That was the key message heard over a couple days recently at GE’s Global Research Center, where the company brought together some 540 control engineers from around GE for its Connected Controls Symposium. For the first time at this annual gathering, GE also included a few members of the industry press, excited to introduce its Industrial Internet Control System—“what we believe will be the next chapter in GE’s digital industrial journey and transformation,” said Jim Walsh, president and CEO for GE Automation & Controls.

While consumer productivity has seen a 19 percent year-on-year rise over the past five years, that number in industry has been only 1 percent. It’s not a very inspiring situation, Walsh conceded. “Although the technologies are available, we haven’t found a way to get the rubber on the road fast enough in industrial,” he said.

Collecting data from controls is key, but it’s typically locked in fragmented, proprietary systems. Getting dynamic, actionable intelligence back into operations to optimize performance and efficiency is even more difficult.

“Customers’ most valuable natural resource is data. But they need to be able to extract out the pearls of wisdom that they need from this data,” Walsh said. “About 1 percent of that data is actually used for anything productive. I think that’s a striking statement.”

Control at the edge

GE’s new Industrial Internet Control System (IICS) is designed to make it as easy as possible for customers to adopt the connected technologies of IIoT to improve operational efficiency—optimizing assets, process performance and productivity. The control system combines Predix (GE’s version of the connected cloud), the digital twin (a digital model of a physical asset that uses data to make continuous adjustments) and edge computing.

“The edge of the Industrial Internet is really controls, no matter how you look at it,” said Rich Carpenter, product general manager, control platforms, for GE Automation & Controls. “We have had to rethink and reimagine what it means to be a controller.”

To start with, the control system comes out of the box with Predix enabled, making it easier to take advantage of connected systems. And since a connected system increased the attack surface, the controller also rebuilds security from the ground up, Carpenter said, with all elements protected independently.

GE’s Field Agent technology is used as the gateway between the asset and the Internet. Three different versions of field agents are available depending on the application, but they’re all geared toward providing secure data collection from the machine. “The field agent is a blind node to the Industrial Internet. You can’t find it; it’s as if it’s not there,” Carpenter said. “It’s a very secure environment from beginning to end, and we will continue to drive that forward.”

The IICS uses mix-and-match I/O, which simplifies scalability, availability and maintenance. And two structural innovations—hypervisor control technology and ComExpress processor technology—“future-proof” the control. The hypervisor control separates operations from software so that the software can be upgraded without affecting the physical controller. The ComExpress processor allows the processor to be updated as needed to support the growing computational power available.

“When we put in a piece of equipment for 20 years, we need to provide incremental value on that every quarter or even more often,” Carpenter said.

Driving outcomes

Historically, controls have been focused on keeping their heads down just to make sure everything keeps running, Walsh said. “That hasn’t changed; they still have to do that. But this controller also has the ability to look up.”

The IICS has not only the deterministic capabilities of traditional controls but also the ability to optimize assets by gathering business data or weather data, for example, to bring optimization concepts back down to the controller, Carpenter said.

As an example, Carpenter referenced a water treatment facility that routinely moves large bodies of water from one location to another. It’s an expensive pump process that uses a lot of energy to move millions of gallons of water. Since it matters little to the process what time of day that water is moved, why not do it when energy prices are at their lowest? With the IICS looking energy spot prices, it can decide what time of day move the water without impacting the process. The utility can save $1.5 million to $2 million a year just from lower energy costs, Carpenter said.

This is a perfect example of why GE is so excited about its new control system, Walsh said. “When you start to think about how you can take data off the asset, apply domain expertise, bring it back to the asset and change the way it runs—that’s when you can really change productivity.”

Applying that concept to the wind power market, which is very dynamic, IICS technology is able to increase profits without raising maintenance costs, said Brent Brunell, technology leader for controls and electronics at GE. In the past, control was based on offline models, with model-based control introduced about five years ago. Wind turbines were delivered, and customers were told to run it in a particular way based on those models.

But tomorrow’s technology will be dynamic optimization based on the data analysis of digital twins, Brunell said, describing the scenario: “If I run this turbine in this market in this way, I can make more money.”

Although GE has long spoken of the “power of one”—the considerable benefits that can be gained from even a 1 percent improvement in efficiency—the IICS’s early adopters have, on average, seen 5-15 percent productivity benefits, according to Carpenter. “That far exceeds the power of 1 percent that we initially envisioned with the Industrial Internet,” he said. “The benefits are real.”

Although customers certainly understand that change is coming to industry, some are a bit leery of the IIoT message. “To a person, they’re tired of hearing the high-level PowerPoint pitch,” Walsh said. “Customers care about outcomes, not necessarily about technology and theory.”

This was a point echoed by Kishore Sundararajan, CTO and vice president of engineering for GE Oil & Gas. “All of the major oil makers want to understand what this Industrial Internet journey entails,” he said during a panel discussion with other vertical leaders. “None of them talk to me about technology.”