GE's software strategy: digitise industry, change the world

Harel Kodesh wants to build a platform where GE and partners can build a new Industrial Internet

General Electric is a company as American as the stars and stripes, baseball, Charlie Brown and the Statue of Liberty. Its roots go back to the 19th century and towering figures such as Thomas Edison, JP Morgan and the Vanderbilts helped shape it. GE helped develop radio and TV and it was intimately involved with the creation of modern power generation, aviation and industrial processes. It is the only company to have been on both the original Dow Jones index and to be there today nearly 120 years later. But that last point begs a question: how can GE prosper in a new age? And one possible answer to that is ‘software and analytics’.

GE CEO Jeff Immelt told his troops last year: “If you went to bed last night as an industrial company, you’re going to wake up this morning as a software and analytics company.” Harel Kodesh, VP and CTO of GE Software, is one of the people charged with keeping GE relevant in an age where binary code often trumps hardware or energy. Israel-born and a former Microsoft, EMC and Amdocs executive, he brings with him a wealth of digital experience and his aim is, by my definition, effectively to reinvent much of what GE does and how it does it.

What is at stake is “a fundamental transformation of the sector”, Kodesh tells me when we meet in London. The industrial ecosystem has to change “in the same way the internet changed data entry and communications”.

Kodesh wants to make GE central to what he calls the “Industrial Internet of Things”, adding digital instrumentation so that industry can become smarter, adaptive, self-healing and capable of generating not just power and product but also insights that create new waves of efficiency and opportunity.

A practical example:

“If you’re a farmer and want to optimise your wheat produce you can seed the field with sensors so you won’t have to irrigate every week,” he says. “The field will tell you when it needs watering.”

Kodesh and the other digirati GE is hiring will bring a very different culture but what will be needed is a synthesis.

“In the jet engine business I always joke that a bunch of people get together every 14 years and you get a new jet engine – modern software is more try, fail and try again,” he says. “In software it’s not bad to release something that fails so long as they fix it; Google Maps was in beta for five years. But you can’t build a jet engine that will fail two per cent of the time. In software, everything is possible - look at Uber and Lyft. You can change the business model very easily. In industry, the model has been set - years ago.”

The difference between consumer software and binary code maintaining aircraft, defence systems and power stations is stark. “Most of our stuff you want to make sure of because [if it doesn’t work] it’s not naked pictures of actresses [being released], it’s taking the power down.”

That said, GE needs to be bold and unafraid of change, he suggests, pointing to the example of Tesla revolutionising the automotive industry. That, Kodesh says, is a warning against complacency and ‘we’ve always done it that way’ thinking:

“If GE sticks to its knitting it might not have knitting anymore.”

Hiring, I suggest, must be a challenge. GE is a great US brand but a new generation of kids see Google and Apple as their leading lights. His answer to this is emphatic and an appeal to the kind of ambition that young people tend to have.

“It’s not about competing but reinventing the world and, with all respect to Google and Apple, this is a new world. We’re going to all the kids and saying here’s the deal: come here for two years - two years is their generation of software - and you’re going to get paid very well and you’re going to do very interesting things and your marketability is going to be high. It’s a calling card.”

But he thinks GE can’t do all this on its own and the company is therefore putting accelerators on the ground in Israel, Singapore, France, the UK and anywhere else where companies might seek to create their own vision of the Industrial Internet. He wants to see a broad array of services running on GE’s Predix platform for developing cloud apps that tie people to industrial machines and generate data that can be mined and analysed.

“Predix is the cloud-based operating system of the Industrial Internet” and a catalyst he wants to “build the tribe” around, he says. M&A is an alternative, I suggest, but he frowns on it because in such a fast-moving space “the company you buy gets disrupted then you change all the time. At this point it’s better to develop partnerships and investments.”

Can a leopard change its spots? Is GE’s culture amenable to this rate of change?

“Microsoft at certain times had much more inertia than GE has,” Kodesh says, recalling his time there when he was senior figure in figuring out how Windows could work in the embedded space and in mobile phones. “They enjoyed the sweet smell of monopoly as did GE. When I was at mobile for six months I had people saying ‘why do you want to do that?’”

Later he left after disagreeing with Microsoft executives on direction.

“Bill Gates was willing to make bets. Steve [Ballmer] wanted to build a better BlackBerry and I came to the conclusion there was no way to square this round peg.”

At GE, Kodesh knows the challenge is enormous but he has something more important to build than a better handset.


Also read:

GE sees robotics, data shaping a new industrial Internet of Things

GE’s Bill Ruh expects human-machine dialogue