As manufacturers put more capabilities into the cloud, one thing is becoming painfully clear: Putting industrial data in the cloud is expensive. The data produced by a turbine, for example, in a single day could cost $20 million a month in storage, according to Rich Carpenter, general manager of control and edge platforms for GE’s Intelligent Platforms division. “The volume of industrial data just dwarfs things that operate at human speed,” he says.
Intelligent Platforms—the segment of GE’s legacy Automation and Controls business that Emerson recently agreed to acquire—has been putting increased emphasis on edge controls and analytics with the understanding that high-fidelity, mission-critical analytics need to be done at the edge rather than in the cloud. That’s not only for cost reasons, but keeping analysis close makes more sense for operations that require quick decisions.
Industry will continue on its journey toward cloud computing, but there will be increasingly more edge computing over time as well, Carpenter says. The trend, therefore, is toward increased compute power at the edge.
Intelligent Platforms has made two product releases recently that harness the combined power of edge controls and analytics to create more actionable intelligence for industry. They do so at two ends of the spectrum—one is an outcome optimizing controller that gives developers more freedom with an open system and the other is a remote monitoring service platform that gives asset managers an easier way to get into the Industrial Internet of Things (IIoT).
Outcome optimizing open controller
When GE released its Industrial Internet Control System (IICS) a couple years ago, the attention was on driving outcomes—with a controller focused not only on the deterministic capabilities that keep everything running, but on a heads-up stance that can gather outside data as well that might influence outcomes. The PACSystems* Rx3i CPE400, the controller behind the IICS, has been leveraging GE’s Predix analytics platform to do that. Now Intelligent Platforms has introduced PACSystems* RX3i CPL410, which provides customers more freedom in their analytics platform and cloud infrastructure.
“The CPL410 is geared toward choice,” Carpenter says. “Some customers want the outcome-optimizing infrastructure but want to use a different cloud.” The latest controller is not tied to the GE cloud at all, he added.
The CPL410 now features PACEdge with Linux technology, enabling customers to not only connect to their preferred analytics-capable cloud service, including Microsoft Azure and Amazon Web Services, but also develop and run data processing Linux-based applications next to the control systems to enhance their processes for better outcomes.
The controller takes advantage of the computing power that’s available with multicore processors—computing power that a traditional distributed control system (DCS) can’t easily make use of, Carpenter notes. The CPL410, like its predecessor, uses hypervisor control technology to virtualize the control layer, providing standard deterministic industrial control in an inner loop and also an optimizing component of that control in an outer loop.
“That came from a big sort of a-ha that we had—when you sit down with your phone, you’re so much smarter than you used to be because you can tap into your phone at any given time,” Carpenter notes. “If you don’t know how to do something, you just pop into your phone and run a video. You have no fear of it because you know you can find out how to do it.”
The same is essentially true for the CPL410. “In my outer loop, if I need more information, I can get it while the inner loop is running,” he explains. “If you’re running a large set of pumps in a treatment facility, you want to pick the time of day when running them is optimal.” The outer loop can be used to check weather, market prices, etc., to make better control decisions. “Do that every single day in a year, and you could save half a million dollars.”
Although future iterations of the platform will be as easy to program as ladder logic or function blocks, Carpenter says, the current version of the open controller is well suited to early adopters who can do some of their own development.
Insight into distributed assets
The other recent launch, Equipment Insight 2018, hits the other end of the spectrum, where customers are looking for considerably more help in getting their processes started. “A large number of customers are a little overwhelmed by the whole IIoT environment. They have a clear idea that they should engage, but they don’t know what to do,” Carpenter says. “The thought of doing their own analytics is beyond a lot of companies. It’s not something they’re staffed to do.”
Equipment Insight was released specifically with that customer base in mind. The software application is used to drive a set of distributed assets, creating dashboards associated with those assets. “It lets you create some alerting or program detection of trends—all kinds of ad hoc troubleshooting tools,” Carpenter says. Although it doesn’t do everything for everybody, he adds, it’s well suited to OEMs or service managers responsible for managing a distributed set of assets. “You don’t need an IT infrastructure and you don’t need to buy additional compute assets.”
With the ability that Equipment Insight offers to enhance performance and processes through centrally managed remote monitoring, OEMs could have an easier time setting up new service revenue models for their businesses and also providing a better experience for their customers. The service enablement platform provides a complete end-to-end solution from data connectivity at the edge to anomaly detection, troubleshooting and case management.
“End customers are now relying on OEMs more and more to keep their equipment up and running,” Carpenter notes. “You just can’t do it if you don’t have that insight.”
If an OEM could typically visit 20 assets in a week, the increased visibility with Equipment Insight could give them the ability to cover 100 assets, Carpenter says, adding that the 20 they’re able to physically visit still are now the ones that actually need maintenance.
Compatible with cloud-based applications and Field Agent technology, Equipment Insight 2018 is well suited to large pumps, backup or portable power generators, compressors and other complex machinery. In a case study with Tecogen, GE shows how the manufacturer of combined heat and power (CHP) systems uses Equipment Insight to manage more than 150 data points coming off each of their systems.
“We’re able to correlate those to one another, to time, back in history, and set up all sorts of trends,” says Joe Gehret, director of field operations for Tecogen.
The remote monitoring and diagnostic capabilities simplify troubleshooting and maintenance to help Tecogen and its customer reduce costs. “That real-time, cloud-based monitoring capability is hugely important to maintain the integrity of our fleet, to make sure our service techs, on their mobile device, get a notification if something goes wrong, get the unit back up and running as quickly as possible,” says Benjamin Locke, Tecogen’s CEO.
Customers can also reduce their costs through peak shaving—reducing the amount of energy purchased from a utility during peak demand hours. “By using the Field Agent, we’re able to actually take advantage of some peak shaving and reduce usage during certain periods of time when the utility calls for it,” notes Dale Desmarais, director of business development for Tecogen.