The internet of things is not what you think

By Steven Butler, The Daily Dose, January 3, 2017

Because the world economy is in the doldrums and could use some help.

The train pulls out of the station as I lean on the throttle — a mile of freight cars in tow — accelerating to over 80 mph. Catching sight of some curves and hills ahead, I throttle back — but, oops! — I’m way over the speed limit. License revoked. Not for real, though, since General Electric is letting me play with its Trip Optimizer at the company’s Global Research Center in Niskayuna, New York. My second try seems auspicious: I hit the brakes, slowing the train but burning way too much fuel in the process.

An experienced conductor would have fared much better, but the fact is that a host of variables — the train’s length, weight and how it’s distributed, as well as the terrain with its hills and valleys, curves and shifting speed limits — are beyond human processing capabilities. And while it may seem like my goal in the virtual conductor’s seat is to have fun, GE’s objective is to make money — by cutting fuel consumption, maintaining a maximum speed and preventing miles of cars from clanging against each other and breaking their couplings.

In homes everywhere, consumers are making the most of cool gadgets like thermostats or electric lights controlled via smartphones. But when it comes to the true Internet of Things revolution, consumer applications will seem a mere curiosity next to the trillions of dollars that could be saved and earned from the still-fledgling industrial internet — which aims to embed sensors in everything and apply vast computing power to optimize performance. According to McKinsey Global Institute, the industrial Internet of Things has the potential to generate over $11 trillion in value annually by 2025 — that’s 60 percent of the current output from the U.S. economy, or roughly equal to everything produced in China.

Which brings us back to the trains. GE developed the optimization software but then had to dumb down the algorithms to accommodate a box small enough to fit into the engineer’s compartment on trains not yet designed to house powerful computers. Still, the company claims up to 10 percent increased efficiency on this industrial-age beast that, industry wide, consumes $11 billion in fuel a year. The real trick, of course, isn’t optimizing one train but rather systems of trains and track usage and cargo, drawing on more computing power and bigger bandwidths. Repeat for airplanes, cars, power grids, wind farms, hospitals, manufacturing plants … and trillions of dollars saved doesn’t seem so outlandish.

A brave new world built on power and efficiency — until you consider its shadow side. “There will be a major disruption in jobs,” says Richard Soley, executive director of the Industrial Internet Consortium, whose corporate members are trying to usher in this new world by setting standards and designing strategies to make it happen. And today’s anxiety over job losses from trade deals could look like child’s play by comparison. Soley points out, for example, that in the developed world, only 6 percent of vehicles are in use at any given time. Between ride-sharing and self-driving cars — and the potential for optimizing thousands of vehicles sharing smart roads — the need to own or drive cars could evaporate, throwing millions out of work and slowing or reversing automobile sales.

And the ramifications don’t end there — imagine the potential impact on everything from insurance companies to traffic cops, even trauma centers. “There’s going to be a shift in the workforce, and nobody can know how it’s going to play out,” says William Ruh, CEO of GE Digital, the company’s recently established software development outpost in Silicon Valley.

GE is trying to position itself at the center of these seismic changes with its Predix software platform (developed in Ruh’s labs), where GE envisions customers, collaborators and competitors will develop apps like consumer apps for Android or iPhone, only more robust for industry. Nick Heymann, analyst at William Blair, says GE stands virtually alone among its industrial competitors in devising a sophisticated digital strategy to capitalize on its deep knowledge of machines. But many doubt that major industrial or digital players like Amazon or Google will risk making themselves so dependent on GE technology. “From GE’s traditional competitors to the hyperscale cloud providers, there are a host of credible challengers,” Paul Miller wrote recently for Forrester Research.

Soley envisions an “ecosystem” of companies playing different parts, and hundreds have joined the 2-year-old IIC, which he cautions must act carefully to avoid antitrust issues. Members are building experimental prototypes to see what works and what doesn’t — from tracking employees on a factory floor, alerting them when they touch a tool they’ve never used to receive on-the-spot training, to wirelessly connecting ambulances and emergency rooms and tracking where drivers spend their time. “It’s really all over the map,” he says.

Yet what strikes Soley and others is just how antiquated the industrial system is — and the striking opportunities that presents for the right technological solution. Offshore oil rigs, explains Soley, must compute drilling adjustments 10 times per second, but round-tripping to big computers in less hostile environments on shore takes a full half second. For decades, manufacturing lines have been run by programmable logic controllers, connected to corporate data systems by the old-fashioned sneakernet — requiring someone to physically walk the downloaded data from one place to another.

But if experts such as Marco Annunziata, GE’s chief economist, are right, then industrial internet solutions could help the world’s economy turn a corner and shift into high gear. All aboard!