For twenty-five years, we've been promised a thoroughly connected world in which our "things" become smarter, safer and save energy. But progress doesn't seem to match the glowing predictions.
The presentation is straightforward and enticing:
Picture this: A 25¢ smart chip inside a light-bulb socket. Networked through the 110V wires, it provides centralized on-off control and monitors the bulb's "health" by constantly measuring electrical resistance. Imagine the benefits in a large office, with thousands, or even tens of thousands of fixtures. Energy is saved as lighting is now under central, constantly adaptable control. Maintenance is easier, pinpointed, less expensive: Bulbs are changed at precisely the right time, just before the filament burns out.
Now, add this magic chip to any and all appliances and visualize the enormity of the economic and ease-of-use benefits. This is no dream. . . we're already working on agreements in energy-conscious Scandinavia.
When did this take place?
There is a one-word giveaway to this otherwise timeless pitch: filament. Incandescent lights have been regulated out of existence, replaced first by CFLs (compact fluorescent lamps — expensive and not so pleasant) and then by LEDs (still expensive, but much nicer).
The pitch, reproduced with a bit of license, took place in 1986. It's from the business plan of a company called Echelon, the brainchild of Mike Markkula, Apple's original angel investor and second CEO.
The idea seemed obvious, inevitable: The relentless physics of Moore’s Law would make chips smaller, more powerful, and less expensive. Connected to a central household brain, these chips would control everything from lightbulbs and door locks to furnaces and stoves. Our lives would be safer and easier. . . and we'd all conserve energy.
The idea expresses itself in variations of the core Home Automation concept, the breadth of which you can visualize by googling "home automation images":
In 1992, Vint Cerf, our beloved Internet pioneer, posed with his famous IP On Everything t-shirt:
This was a modern, ringing restatement of Echelon's vision: The objects in our homes and offices will have sensors and actuators. . . and a two-way connection to the Internet, to a world of data, applications, people (and, inevitably, marketing trolls).
It's been a quarter century since Echelon started, more than two decades since Vint Cerf's pithy yet profound prophecy. We now speak of the Internet Of Things and make bold predictions of billions of interconnected devices.
Earlier this year, Cisco invited us to "capture our share" of the $14.4T (yes, T as in trillion) business opportunity that The Internet of Everything (IoE) will create in the coming decade. Dave Evans, Cisco's chief futurist, tells us that within ten years we'll see "50 billion connected things in the world, with trillions of connections among them".
Maybe. . . but that's a lot of "things".
As Network World points out, "[m]ore than 99 percent of physical objects are not now connected to the Internet". The exact percentage matters less than the existential truth that the obvious, attractive, inevitable idea of a universe of interconnected objects is taking a long, long time to materialize.
Does the concept need a Steve Jobs to coalesce the disparate components into a coherent, vibrant genre? Are important pieces still missing? Or, like Artificial Intelligence (rebranded as Machine Learning in an attempt to soothe the pain of repeated disappointments), are we looking at an ever-receding horizon?
Echelon's current state (the company went public in 1998) serves as a poster child for the gulf between the $14.4T vision and today's reality.
First, some context: Mike Markkula, who is still Vice Chairman of Echelon, has assembled a strong Board of Valley veterans who have relevant experience (I know several of them well — these aren't just "decorative directors"). The company's Investor Overview offers an impressive Corporate Profile [emphasis mine]:
"Echelon Corporation is an energy control networking company, with the world's most widely deployed proven, open standard, multi-application platform, selling complete systems and embedded sub-systems for smart grid, smart city and smart building applications. Our platform is embedded in more than 100 million devices, 35 million homes, and 300,000 buildings and powers energy savings applications for smart grids, smart cities and smart buildings. We help our customers reduce operational costs, enhance satisfaction and safety, grow revenues and prepare for a dynamic future."
But the latest Earnings Call Presentation paints a different picture:
The Gross Margin is good (58.5%), as is the company's cash position ($56.7M). . . but Echelon's business is a tiny $18M — about a millionth of Cisco's predicted motherlode. That's a decrease of 38% compared to the same quarter last year.
So, we have a company that's in the hands of competent technologist who have deep knowledge of the domain; a company with real, proven products that have been deployed in millions of homes and offices— but with little revenue to show for its technology and experience.
This seems to be the case for the Everything Connected industry in general. There's no General Electric, no Microsoft, no Google (the latter abandoned its PowerMeter initiative in 2011).
Why not? The answer might lie in the Echelon presentation already mentioned:
After more than 25 years of developing devices and platforms, Echelon concludes that the Internet of Things isn't going to be felt as a direct, personal experience. Instead, it will be mostly invisible: components and subsystems in factories, warehouses, fleets of trucks and buses, office buildings. . .
Consumers certainly don't have to be sold on the benefits of connected devices. We can't function without our smartphones, tablets, and PCs. But once we stray outside the really personal computer domain, the desirability of connected devices drops dramatically.
The dream of giving sensors, actuators, and an Internet connection to everyday objects feels good, until one looks at matters of practical and commercial implementation. Will the software in my smart toaster be subject to a licensing agreement? Will it stop toasting if I don't renew my subscription? (This isn't just a dystopian strawman; one electric car manufacturer says it can remotely disable the battery if you don't pay up.)
And then there are the (very real) security and privacy concerns. Could our appliances be hacked? Could my toaster spy on me, collect more data to be used to peddle related goods?
Home automation and security systems seem like a natural fit for the Internet of Things, but they're still expensive, complicated, and fragile - if not hopelessly primitive. Some connected thermostats, such as the Nest (with its smoke and carbon monoxide detector), work well, but most of them are stubbornly user-hostile.
When we wander into the realm of connected appliances what we see are novelties, fit only for hobbyists and technofetishists (do we really need a toaster that sends a tweet when it's done?). This is nothing like the smartphone wave, for a simple reason: Appliances are just that, appliances. It's word we use as an insult to describe a boring car.