General Electric: Post-industrial revolution
The shift to intelligent devices exposes group to stiff competition from tech companies
General Electric’s latest technology for boosting manufacturing productivity is a gadget that feels like it would be more at home in an amusement arcade or theme park than in an industrial laboratory. At the group’s global research centre in Niskayuna, New York, you can enjoy the sensation of flying around its steam turbine plant in Schenectady, six miles away.
A virtual reality 3D scan of the plant, accurate to every last machine tool and piece of pipework, is projected on to a screen and special glasses. Using a PlayStation-style controller, you can move around the factory floor, swooping over assembly lines or hovering over a stack of components.
The system is a lot of fun to use, but the intent behind it is entirely practical: to allow GE’s engineers to work on design, layout and workflow at its plants without actually being there. If they want to move in a piece of new equipment, for example, they can use the simulation to see if it will fit. The systems for collecting and analysing manufacturing data — fitted in at least 75 of GE’s 590 plus factories — can cut costs by 15 per cent or more, say company executives.
The technology is just one part of a radical overhaul designed to transform the 123-year-old group into what Jeff Immelt, chief executive since September 2001, calls a “digital industrial” company. At its core is a drive to use advances in sensors, communications and data analytics to improve performance both for itself and its customers.
GE’s products such as aero engines, power generation equipment, locomotives and medical scanners are being made part of the “internet of things” — intelligent connected devices that can transmit information and receive instructions — and the company is building new capabilities in software to understand and manage those machines.
The potential rewards for success are enormous for a group that has sold off large chunks of its business to refocus on its industrial operations.
“It is a major change, not only in the products, but also in the way the company operates,” says Michael Porter of Harvard Business School. “This really is going to be a game-changer for GE.”
If it fails, however, it could prove a decisive factor in hastening a further break-up of the group.
The power of the industrial internet of things is only just starting to be explored. Companies in areas from manufacturing to energy, transport, and mining collect huge volumes of data, but use only a fraction of it.
If GE can succeed in finding ways to use that information to cut its costs and raise the productivity of its products and services, it could gain a critical competitive advantage over rivals such as Siemens, Mitsubishi, United Technologies and Rolls-Royce. It also aims to create an important new source of income from cutting costs and boosting productivity for other companies, even if they are not using GE equipment.
Entering this world, however, is also bringing GE new competition. In setting itself up as a software business that can help other industrial groups reap the benefits of the internet of things, GE will be taking on Microsoft, Amazon, IBM, Oracle and SAP.
“GE is drag-racing with the best technology companies in the world,” says Frank Gillett of Forrester Research. “Kudos to them for trying, but I think they will find it harder than they think.”
Getting out of finance
In April, GE embarked on one of the most radical changes in its history, saying it would sell about 90 per cent of GE Capital, the financial services unit that just a couple of years ago provided about half the group’s $14bn earnings.
That decision was welcomed by analysts and investors — its shares, after lagging behind the wider market for most of Mr Immelt’s tenure, have risen 17 per cent in the past 12 months. The crisis of 2007—09, which forced GE to cut its dividend and lose its triple-A credit rating, convinced many of the business’s potential for disaster.
The regulation that followed, placing additional burdens on GE as a “systemically important financial institution”, meant that even in good times the returns on capital looked unattractive.
Yet while financial services might have been a dangerous artificial stimulant, there is a big question about how fast GE, can grow without them.
When Jack Welch, Mr Immelt’s predecessor, took over GE in 1981, it was viewed by many as a “GDP company”: the kind of big, boring business that grows only as fast as the gross domestic product.
To avoid that tag Mr Welch pushed into financial services and other businesses that could grow more rapidly. Without that business, GE faces the prospect of returning to GDP rates of growth.
With a market capitalisation of $250bn GE is one of the most valuable brands in the world. The group employs over 300,000 people and its most profitable sectors last year were power equipment and aviation which both contributed about $5.5bn.
The sale of most of GE’s financial services businesses means that reported earnings per share are expected to have dropped 21 per cent last year. In 2016, the slowdown in the world economy and the fall in commodity prices that has hit GE’s businesses providing equipment for oil production and mining will be a drag on growth.
Brian Langenberg, an industrial analyst who chairs the business school at Aurora University in Illinois, expects GE’s organic sales growth from now on to be in line with global GDP, forecast by the International Monetary Fund to be 3.6 per cent this year.
The industrial internet, however, offers GE a hope of escaping that uninspiring prospect. The falling cost and rising power of sensors, communications devices and data processing mean that just about any product can be made capable of sending and receiving information and commands. The internet of things is best known today for consumer applications such as fridges that can order your groceries. The more important applications, however, are likely to be in industry.
By 2025 the economic benefits of the internet of things could be $11.1tn a year, according to the McKinsey Global Institute. It estimates that about 40 per cent of those benefits could come in factories and work sites such as oilfields, with transport and urban infrastructure accounting for a further 30 per cent.
Marco Annunziata, GE’s chief economist, believes that as companies work out how to exploit the potential of the new technologies, it could unleash a new productivity revolution in industry. Since 2010 productivity in US manufacturing has stagnated, rising just 1 per cent a year compared with the annual 4 per cent growth in the previous two decades. Mr Annunziata says the industrial internet could help bring back those higher rates of growth.
“The first ICT revolution came in the 1990s: using computers as ways to gather and organise information. Now we are literally making machines more intelligent,” he says. “This tying together of the digital and the physical is something we have never seen before.”
He cites mining companies, under huge pressure to cut costs because of weak commodity prices, as an example of businesses that are keenly interested in the potential of the technology.
They already record large amounts of data — Teck Resources of Canada says it has 200 sensors on every mining truck — and if they can analyse it properly they can discover ways to improve efficiency, such as predicting more precisely when failing parts need to be replaced, reducing the time when expensive machinery sits idle.GE says that at one of its mining customers, trucks that were previously available for use 70 per cent of the time are now available 85 per cent.
Jim Heppelmann, chief executive of PTC, a software company that works with GE and other manufacturers, says that sort of boost to productivity represents a critical competitive advantage.
“If you have a 10 to 20 per cent cost advantage on a product with 3 to 5 per cent margins, you’re going to walk all over the competition,” he says.
GE executives believe it can be in the vanguard of this revolution. The company is spending $1bn a year to boost its digital capabilities, hiring 1,000 software engineers and data scientists and setting up a new data analytics centre in San Ramon, California, just across San Francisco Bay from Silicon Valley.
In the next month, it is expected to launch Predix, its software platform for managing and analysing industrial data.
“In industry, there is a lot of stuff that people don’t know how to find and don’t know how it is performing,” says Beth Comstock, who leads new business development at GE. “We can analyse how it is performing, [and] we can predict what will happen.”
That analysis will be applied to both GE products and to those made by other companies. A modern locomotive is a “rolling data centre”, as Mr Immelt puts it. By analysing that data and cross-referencing with its rail traffic management system, and sending instructions to trains, GE can squeeze out an extra mile per hour of speed for US railway operators worth $200m per year in extra profits.
Similarly, Toshiba is working on a pilot project with GE to develop an application for installing, operating and maintaining lifts.
GE expects the use of Predix to grow rapidly, with 500,000 products under management by next year. Ms Comstock argues that many customers will want GE to provide them with an integrated package of products and services to deliver specified outcomes, with data analysis a central part of the package.
So far GE is the industrial company that has made the strongest commitment to the industrial internet, although other manufacturers such as Bosch of Germany and France’s Schneider Electric have also been starting to explore it. Siemens, GE’s main European rival, has announced only tentative initiatives.
Friends and family
GE accepts it is facing stiff competition in data analysis from specialised software companies. It argues, however, that rivals start from a position of weakness because they do not have the same understanding of industrial machines.
“The domain knowledge for this is hard to come by. It is held by a small number of people,” says Bill Ruh, head of GE’s global software operations. “We are the best of the best, because of what we’ve done in combining physics and analytics.”
For all Mr Ruh’s confidence, grafting a world-class software business on to an industrial conglomerate is not easy.
In its attempt to rebrand itself as “the digital company that’s also an industrial company”, GE has been running adverts showing a young software engineer trying to explain his new job at the company to family and friends.
In one, his friends are unimpressed by what he does for power plants and hospitals, but enthuse over another engineer who is working on “the app where you put fruit hats on animals”.
It is funny because it is true: the gulf between the cultures of Palo Alto, California and GE’s headquarters in Fairfield, Connecticut is wide.
“They should not underestimate the challenge of reinventing themselves as a digital company,” Mr Heppelmann says. “Silicon Valley is a very special place in terms of its culture, its star system, its remuneration. That’s something GE can’t bring.”
It is not just in terms of recruitment that Silicon Valley poses a threat to GE.
“GE is in a race to capture customers before the likes of Amazon get better at meeting industrial requirements, and before customers get comfortable about using them,” says Mr Gillett.
Establishing General Electric as a leading company in the industrial internet would be a crowning achievement for Mr Immelt, who for much of his time at the top of GE has laboured in the shadow of Mr Welch. If the strategy fails, however, it will be his failure.
The arrival of its first activist investor — Nelson Peltz’s Trian Fund Management last year revealed a stake of just under 1 per cent — could also prove problematic. While broadly supportive of Mr Immelt, Trian has urged him to do more to raise profit margins, and it is clear that its amicable tone could sour.
If the costly bet on the industrial internet fails, then so will GE’s dreams of achieving growth through technological leadership. A less glamorous future of deeper cost cuts, lower levels of investment and perhaps a further break-up would await.