Cutting Decades Out of the Optimization Curve

GE’s goal of transforming itself from an industrial company into a software and analytics company becomes clearer as the company positions itself around the new Predix System.

“There’s a lot of sensor technology already in place across industry,” said Steven Martin, chief digital officer at GE Energy Connections. “What we’re trying to do is reduce the time it takes [for an operation] to get to an efficient state.”

Explaining his statement, Martin said that end users have been optimizing equipment in place the same way for years using basic operating data gathered from sensors and controllers. However, if you put more sensors in these assets to extract more data for greater levels of analysis and insight with integrated applications, you can get the yield up on those assets very quickly, he said. “As we’ve seen with applications in the renewable energy industry, you can cut decades out of optimization curve.”

It’s around this vision of speeding optimization and remaking industries across supply chains that GE sees its future. Over the past few years, GE has been on a path to realign its business as a software company enabling the Industrial Internet (GE’s preferred term for the Industrial Internet of Things). It's much the same path that many major automation technology suppliers are on today, but GE’s announcements at Minds + Machines 2016 appear to position the company well ahead on its path to offering a complete package of IIoT-enabling software for industrial end users and OEMs.

GE reports that orders from its software portfolio this year are on track to climb 25 percent to more than $7 billion. The company says this rate of growth makes “GE the fastest growing digital industrial company in the world.”

Predix Becomes a System
The core focus of this year’s Minds + Machines event was around the release of the Predix System, which makes GE’s Predix a distributed system. GE calls the Predix System a “complete edge-to-cloud offering.” According to the company, the Predix System is “a software architecture and services designed to make any machine an intelligent asset and to bring visibility, control and analytic insights to every part of industrial infrastructure and operations.” The Predix System can run on a variety of operating systems, devices and form factors, including sensors, controllers, gateways, servers and the cloud.

In his keynote address, Jeff Immelt, CEO of GE, said that since the company opened the platform to customers earlier this year, 19,000 developers are currently building on Predix. He expects to have 20,000 developers working on the platform by year-end.

A key aspect of the Predix System is its enablement of digital twins (also known as virtualization) for predicting and optimizing the performance of machines. The reason the digital twin concept is so integral to the Predix System is that it is intended for use across all assets in a business. This level of application takes the digital twin from its primary use in predictive maintenance applications into machine learning for optimization of the entire business and its operations. 

Underlying the criticality of machine learning to GE’s digital twin concept are the acquisitions of Bit Stew Systems and Wise.io. Both companies will be leveraged for the data intelligence and machine learning capabilities they can bring to the Predix System. Read more about these acquisitions.

Adapting to New Business Models
Pointing out that GE’s strategy around the Predix System has focused the company on the edge computing and automation layers of technology, Jim Walsh, president and CEO of GE Automation and Controls, admits that this technology concentration is much the same with GE’s competitors. “What’s different,” he said, “is that there isn’t anybody who has put together the pieces organically like GE is doing in terms of partnerships from the edge to the cloud.”

Walsh added an important caveat to this statement: “When the center of your universe is data, that brings in non-traditional competitors.”

In other words, as much as IIoT is changing the landscape for end users and OEMs, it’s doing the same to technology suppliers.

In our discussion about how technology is changing the competitive landscape for both manufacturers and suppliers, I asked Walsh about GE’s growing ecosystem and the clear overlaps in GE’s technology with some of its partners’ technology. “We’ve evolved in our thinking,” he said. “Two years ago we had lots of discussion about technology overlaps. What we’ve learned is that there will be some overlap and you've got to get comfortable with it. We can't build it all ourselves. Data is the currency and whoever controls that and has access to it will determine who are the winners. GE is in unique position as an OEM to have ownership of the data [for its products] and that makes it less threatening.”

A bigger change for GE is transforming the mindset of its own OEM customers. “We’ve always been good at driving competitiveness internally,” Walsh said. “The big change now for us is delivering this [capability] externally to help OEM customers position their business in terms of delivering outcomes to their customers rather than selling pieces of equipment. They have to start asking themselves: What are the outcomes that are impactful to our customers’ businesses. It’s not the traditional approach [of just selling equipment] anymore.”

In the industrial OEM space, this transition is slow in coming as most OEMs still focus on selling equipment as they’ve always done rather than selling what their equipment provides as a service.

Rich Carpenter, general manager of control platforms at GE Automation and Control, pointed to the example of Schindler Elevators, which is transforming itself around this new model. “Their sales are down 5 percent this year,” he said, “but their profit margin is up because they’re providing better service” and selling on that model.

Edge computing is a key aspect of this new equipment-as-a-service business model, as well as with production optimization in general. After opening the Predix System to outside developers this year, Walsh noted that GE has rapidly been shifting its focus to concentrate more on edge computing. It became clear—through a combination of bandwidth issues and business desires to keep critical data on premise—that use of edge computing for analysis is a more sensible approach than uploading increasing amounts of data to the cloud, he said.

With all the announcements made at Minds + Machines 2016, the massive scope of the Predix System and the ecosystem GE is developing around it seem to position Predix predominantly as a technology for the biggest companies. So I asked Walsh about GE’s strategy around Predix as it pertains to the vast majority of manufacturers, which are small to mid-sized companies.

“You have the ability to flex with software,” Walsh said, “to sell it as a service or through licensing. You have to make it easy for people to get started, which means componentizing it.” As an example of how GE enables this, he referenced GE’s Field Agent, which links existing equipment to Predix for collection and analysis of data. “This is a sub $1,000 investment that lets you get started to show how analytics can lead to new results. If our pitch was that you had to start with all of it, that would be a non-starter for most companies.”

Companies in this article
More in Control