Subscribe to our newsletter
When did shopping online become more like driving a 200-mile-per-hour racecar? Quite recently, thanks to something called “digital twin” technology. Now it’s going to change railroads, airlines, factories and the rest of the business world.
Winning a Formula 1 race is no longer just about building the fastest car, hiring the bravest driver and praying for luck. These days, when a McLaren team races in Monaco or Singapore, it beams data from hundreds of sensors wired in the car to Woking, England. There, analysts study that data and use complex computer models to relay optimal race strategies back to the driver.
Meanwhile, the most advanced retailers don’t just advertise based on broad demographic information such as age and income anymore. They develop psychographic models of customers. Do you read Michael Lewis books and enjoy independent films? Do you prefer cycling to golf, celebrate your wedding anniversary in May and live two hours from your parents? Sophisticated retailers pour such data into models that predict your tastes. When you shop online, they’ll carefully select ads you’ll find more seductive.
What’s the connection between racers and online shoppers? To GE Vice President of Software Research Colin Parris, it is clear. The McLaren race crew and the online retailers both harness data and use algorithms to make reasonable projections about the future, Parris explains. The concept is called digital twin. “The opportunities of the digital twin are huge,” he says.
The idea, which involves building a digital model, or twin, of every GE machine, from a jet engine to a locomotive, will grow and create new business and service models through the Industrial Internet. Parris lays out how these digital twins can then be analyzed: A jet engine that would normally be overhauled every 24 to 36 months, for example, may not require such a service until after 38 months based on data from its digital twin. It’s an approach also being embraced by the U.S. Air Force.
Parris admits he is fascinated by Google, Apple and Amazon. The three companies relentlessly gather psychographic data and use analytics to predict what customers want. Their efforts helped them generate $338 billion in combined revenues in 2014, up 10 percent from 2013. (The average annual revenue growth for S&P 500 companies is less than 3 percent.) Parris is convinced that their ability to predicting their customers’ behavior has helped this growth. He says the Industrial Internet is now at a tipping point, promising a similar opportunity for companies such as GE.
“We are doing the same thing with digital twin. We are getting all the data we can possibly get about our engines — data for every flight, of the physics of the engine blades and how the engine is operating, data about ambient temperature and dust levels — and then I can predict exactly when to bring the aircraft in for inspection,” Parris says.
Parris, who joined from IBM a year ago, has been spreading the work of the digital twin throughout GE, from health care to aerospace. In doing so, he’s demonstrating the advantage of the GE Store — the sharing of ideas and expertise between GE’s various businesses. GE is already using digital twins that can optimize and help design wind farms and power plant-scale circuit breakers. He will discuss his vision at GE’s fourth annual Minds + Machines conference about the potential of the Industrial Internet, to be held Sept. 29-Oct. 1 in San Francisco.
For Parris, the digital twin opportunities for GE start with its big machines — aircraft engines, locomotives and gas turbines. Using the digital twin approach, a GE service team will know not only when to bring an aircraft in for inspection but what parts to have on hand and how long the jet will be out of service.
Previously, maintenance was done based on averages derived from field experience. Using this new approach, a locomotive train might not undergo a replacement of its bearings if they still looked good upon inspection and the locomotive’s digital twin signaled they didn’t need replacement. Conversely, the twin might tell inspectors that the bearings in another locomotive car that endured heavier use or harsher temperatures should be replaced early to avoid failure.
It’s a win-win for GE and its customers. For customers it means less downtime for everything from ships to power plants, where the technology can optimize fuel economy and reduce unplanned downtime. For GE, says Parris, it will boost service-contract profit margins.
“The twin is a collection of algorithms and models that give us continuous insight,” Parris says. “We have the inspection data to understand damage to a machine, we have the digital and physical models that can predict the damage and we have analytics that actually work for industrial problems.”
GE’s chief executive, Jeff Immelt, has brought these capabilities together into a new unit called GE Digital, combining the firm’s IT capabilities, including software and analytics. The unit is expected to generate $6 billion in revenues this year and is forecast to become a top 10 software company by 2020, up from 15th currently.
Parris says the digital twin approach will also help GE develop new areas of business, perhaps developing software to schedule the use of hospital beds and imaging equipment such as MRIs.
GE, he says, could expand into new areas and build new revenue streams from such things as customizing optimizers and dynamic controls. For example, Parris says, GE could sell algorithms and services to other firms that need to gather and analyze large amounts of data. “Once we have built this,” Parris says, “we can take these capabilities to other industries.”
“We will be able to influence a lot more companies,” he explains, “because we will manage the system of systems.”