by Rob Preston, February 2017
FedEx founder Fred Smith famously observed decades ago that the information the company collects about the packages it delivers is as important as the packages themselves. In today’s digital economy, manufacturers and service providers across industries are taking a cue from the FedEx visionary, as they start to value the information they collect from their wind turbines, railroad tracks, farmlands, medical systems, lighting fixtures, automobiles, oil fields, mine shafts, appliances, jet engines, and other internet-connected “things” as much as those things themselves. That data, often combined with third-party data and analyzed to reveal new insights and market opportunities, is even creating wholly new business models. And it’s helping businesses realize the grand promise of the Internet of Things (IoT).
No blue chip company has embraced this digital worldview more aggressively than GE, which is positioning its cloud-based Predix operating system as the platform for building applications that connect to its and other companies’ industrial assets. Applications built atop Predix, developed on Cloud Foundry’s open source technology by the company’s GE Digital unit, collect machine-grade data at scale and analyze it to help customers improve system uptime, cut energy costs, and deliver a variety of other “outcomes.”
Having sold off most of its financial services businesses in recent years, GE, under the leadership of CEO Jeffrey Immelt, is doubling down on its aviation, power, transportation, oil and gas, healthcare, renewable energy, and other industrial operating units, betting its future on turning them into “digital industrial” companies. That isn’t marketing bravado. GE Digital, a San Ramon, California–based operating unit the company launched in September 2015, already generates about US$6 billion in revenue from the software and digital services it packages with GE’s industrial systems.
Led by 30-year tech industry veteran and GE chief digital officer Bill Ruh, GE Digital has hired more than 2,000 software developers and engineers to complement the domain experts in GE’s various industrial divisions. It has also signed up more than 270 integration, technology, and other corporate partners, including Oracle, Accenture, Tata Consultancy Services, PwC, and Intel.
One of GE Digital’s biggest market opportunities for Predix is in helping customers with asset performance management (APM)—predictive system maintenance and improved system reliability. It’s a business augmented by GE Digital’s September 2016 acquisition of APM software developer Meridium.
For evidence that GE is truly committed to its “digital industrial” vision, look no further than Current by GE, an energy services company it set up in October 2015 that combines GE’s LED lighting, solar, and other energy products, with the Predix software platform at the center.
Instead of just selling customers LED lighting fixtures to replace their inefficient incandescents and fluorescents, for example, Current will also up-sell a package of hardware and software tools that helps customers further reduce their energy costs.
As a straight replacement for incandescent and fluorescent bulbs, LED lighting can cut a commercial or industrial company’s energy usage by as much as 50 percent, quickly paying for itself. The next step for customers is to add sensors to the LED lighting—which, because it’s a digital platform, comes with its own processing and storage capabilities, notes Current’s chief technology officer, David Bartlett.
As such, the sensors embedded in the lighting—which is interconnected across a building, campus, or municipal environment via wireless or wired networks—can gather data on environmental conditions and movement. Analyzing that data can thus cut customers’ energy costs by an additional 10 to 20 percent by doing things such as harvesting daylight or adjusting the lighting based on occupancy—much more precisely than with motion detection, Bartlett says. One study estimates that commercial/industrial environments have been overlit by three to four times in the last 30 or 40 years, he says.
Current’s LED lighting sensors can connect to any energy-using equipment—sometimes through partnerships with other companies, or through Current’s recent acquisition of Daintree Networks—to feed data on heat, airflow, noise, and other conditions. The sensors can even record video and audio. Sensor-laden LED streetlights, for instance, can not only help cities cut their energy costs, but the visual data they collect can also help city managers ease traffic congestion and plow the snow faster, as well as help residents find the nearest parking spots.
“There’s a lot more real-time decision-making here,” Bartlett says. “That’s where Predix becomes so important, because that was built for this level and amount of data collection and analytics, with the right security.” Bartlett compares the multi-function LED lighting platform to the smartphone, which is as much a camera, email terminal, calendar, recorder, and alarm clock as it is a voice communication device. “It’s to the point where I don’t even think of it as lighting first and foremost,” he says. “I think of it as a digitalization medium that oh, by the way, also provides light.”
Current’s customers include Walgreens, McDonald’s, Hilton Worldwide, JPMorgan Chase, Simon Property Group, and Intel.
In business terms, GE executives describe the goal of APM as “no unplanned downtime.” For example, a GE jet engine can generate terabytes of data every flight. Collecting and analyzing that data helps the maintenance crews of its airline customers determine when to replace a component before it malfunctions, avoiding costly unscheduled downtime. To put that opportunity into context, Ruh notes that 41 percent of the times a flight is canceled or delayed it’s because of a mechanical failure. How much would it be worth to an airline and its busy customers to eliminate such failures?
Another digital opportunity for GE is in improving system performance. For example, to improve the efficiency of its wind turbines, GE used to have to develop a new turbine blade design or a new material. Now the company can use analytics to fine-tune the position of those blades based on wind conditions. “The idea that you can do that through software was unbelievable at one point, but now it’s becoming the mantra for many folks,” Ruh says. “They know now that if they have a way to consistently add an upgrade and get more performance, they’ll learn more as they get better data, as they get smarter on analytics.”
GE Digital is also getting into schedule optimization. One application now written on Predix, called Trip Optimizer, combines GPS data with railroad track profile, rail condition, weather, and other data, letting railroad companies plot the best time and route for each trip, adjusted in real time as conditions change. Trip Optimizer, installed on more than 5,000 passenger and freight trains, has saved rail companies more than 90 million gallons of diesel fuel since it was introduced a few years ago, according to GE Digital estimates.
Yet another big opportunity is in field service management, equipping field personnel with Bluetooth-enabled measurement tools that can read industrial-machine data without someone having to climb a pole or slide under a railway car and then shout numbers back to a coworker. It’s a capability that will be augmented by GE Digital’s pending acquisition of ServiceMax, a developer of cloud-based inventory and workforce software.
Automating data collection not only cuts the number of maintenance people on a job from two to one and reduces errors, Ruh says, but it also reduces the time it takes to do measurements by 70 to 80 percent. “The person has the ability to fundamentally change what he’s doing based on the data instead of having to make two or three trips,” he says. “This is going to change how you service and optimize anything in the world.”
GE Digital and its partners are rolling out Predix-based applications and services in three overlapping stages. The first is internal. At the direction of Immelt, Ruh’s team is working with colleagues in the company’s various industrial units to analyze data and digitize processes with the mandate of squeezing out US$500 million in productivity improvements this year—and a total of US$1billion by 2020. Part of those savings goes to the GE bottom line, while the rest goes back to GE Digital to invest in that business.
“When a customer wants to buy Predix from us, they don’t want to go see another customer,” Ruh says. “They want to go see GE Aviation, GE Power. Why? Because they want to learn how to do it themselves. I want to be clear: This is not a technology sale. What we’re selling are outcomes.”
At the same time it’s proving its digital capabilities internally, GE Digital is also engaged in the second stage: plugging Predix-based applications into the parent company’s installed base of power, transportation, healthcare, and other industrial systems. With GE’s wind turbines, for example, GE Digital engineers can customize the company’s hardware-software PowerUp platform to increase a wind farm’s output by up to 5 percent. “We come in and model your wind turbines and we figure out if they are being the most efficient given the wind that’s produced in that area,” Ruh says. “Then we can figure out for ourselves how much more output we can guarantee you. And then we come in and say, ‘We’ll deliver this outcome,’ and then we share on the benefit side.”
I want to be clear: this is not a technology sale. What we’re selling are outcomes.
The third stage involves applying GE Digital software and services in industrial settings relatively new to the parent company, beyond areas in which GE builds its own products. These new markets include managing commercial and residential building systems, automotive production lines, and consumer goods inventories. For example, Switzerland’s Schindler Holding is tapping Predix to analyze data generated from its connected elevators, escalators, and other building systems to resolve service problems before they happen, reducing downtime. Japan’s Lixil, best known in the US for its American Standard brand, is piloting Predix to optimize schedules for installers of its bathroom fixtures, with the goal of reducing costs and minimizing project delays for its builder customers.
GE leaders estimate that the data collected from the sensors embedded in its and other companies’ machines will increase 100-fold by 2020, fueling a data analytics and digital services business that could generate US$15 billion in incremental revenue for the company by then. The overall “industrial internet” market—GE’s term for its slice of IoT—will reach US$225 billion by 2020, the company predicts.
Ruh emphasizes that Predix is an open platform, and that GE Digital won’t become the be-all-and-end-all software and digital services provider for its industrial customers. Since making Predix available in February 2016, the unit has signed up more than 21,000 developers to build applications on the cloud platform. Among its 270 corporate partnerships, GE Digital is working with Oracle to marry its operational technology (OT) (the software that collects, organizes, and analyzes industrial machine data) with Oracle’s information technology (the enterprise resource planning [ERP] and other software that manages companies’ manufacturing, supply chain, financial, HR, and other business processes).
GE Digital and Oracle are training and certifying other partners to integrate and implement Predix solutions with Oracle Cloud applications. For example, a factory might want to combine uptime data pulled from a production-line machine with Oracle Enterprise Resource Planning application data to make its forecasts more reliable. For GE to be able to sell an aforementioned “outcome”—especially if it’s a financial result, such as higher profits or sales—this OT-IT connection will be crucial.
Dave Bartlett, chief technology officer at Current, an energy services company GE set up in October 2015 (see “Lighting as Ubiquitous Digital Platform” sidebar), singles out Oracle in discussing the massive opportunities ahead for Current as it integrates the energy-related data it’s collecting and analyzing with data from a range of third-party sources.
“Oracle’s huge value to the world is the amount of installed systems it has and information it’s gathering about the way people run businesses, including output, productivity, maintenance, ERP—across industry verticals,” Bartlett says. “Imagine fusing that knowledge, that data with the type of data that Current is pulling in real time from building environments, and imagine using analytics that exist in GE and Current with the analytics and capabilities that Oracle has. That’s huge. I mean, what two other companies could do that?”
GE Digital’s Ruh emphasizes that GE’s customers are only starting to understand the central role that data can play in improving their industrial operations and output. “The idea has to be that you’re going to use our capability for the long haul to drive productivity and efficiency,” he says. “And productivity and efficiency aren’t going to come through a magic bullet, by this one thing we’ve done. It’s a continuous improvement of algorithms—some that we developed, some that our partners developed, some that customers developed, and some that they’ve let us develop for them—that continuously adds value of small percentages, adding up to a big number.”
For additional insight into GE Digital’s ambitious strategy, read the Oracle Voice article on Forbes.
Rob Preston is editorial director in Oracle’s Content Central organization. Follow Preston at @robpreston.