CMU Grad Students' Universal Sensor Could Make The Internet Of Things Cost Less And Do More

Jul 26, 2017

John Essey and his wife Adrienne recently sold their house in Dormont and moved to Allison Park to be closer to family.

Essey said his “excessive” collection of sensors in his Dormont home was both a selling point and a drawback for potential buyers.

“They were impressed, and it was nice for someone who was tech minded, but some people were still kind of put off as far as why all these sensors were in your house? Do you have cameras? What’s the purpose of these things?” Essey said.

John Essey in his Allison Park home.
Credit Liz Reid

Essey is an engineer at Uber and an early adopter of the internet of things, or IoT, which is when everyday objects, like doorbells and thermostats, are connected to the internet. Essey’s house knows when he or Adrienne pulls onto their street because of little digital fobs they have on their keychains.

“Doors unlock, kind of sets the mood for the rest of the house, too, turns on lights, sets the thermostat accordingly,” he said.

He can control his lights with his Amazon Echo or an array of touchpad sensors he’s installed throughout the home.

Essey’s setup might sound a little like science fiction, but it’s a prototype for a future experts say isn't far away as IoT technology becomes ever more ingrained in daily life.

“There are two ways people achieve the IoT, especially if they want to turn an environment into a smart environment,” said Gierad Laput, a PhD student at CMU’s Human Computer Interaction Institute.

He said one way is to buy a bunch of smart appliances and devices.

“Obviously those cost money, and then at the same time they often don’t work together because they’re made by different manufacturers,” Laput said.

The other way is to get sensors, and put them on everything you want to monitor.

“But then those get really unwieldy, and you’ve got all these things sticking around and they look ugly and socially obtrusive,” Laput said.

So Laput and his team wanted to see if they could build one sensor that could monitor a whole range of activity in a room. It doesn’t look like much -- just a 2-inch square circuit board that plugs into the wall. The board senses about a dozen different facets of its environment: vibrations, sounds, light colors and so on. The sensor communicates wirelessly with a computer, which interprets everything the sensor picks up.

Laput demonstrated how the sensor works by turning on a blender. Immediately, a computer screen displayed a visual representation of the vibrations given off by the blender, and words saying “blender on” came on the screen.

Laput turned on a light, and the screen said, “light on.”

Laput said he imagines both domestic and commercial applications for such a sensor. It could tell buyers they left the stove on or that the bathroom at their family-owned restaurant is almost out of paper towels.

Kevin Ashton, who literally invented the term the internet of things, said these are the kinds of problems that don’t make sense to solve using IoT components until these components become really cheap.

A computer screen displays all activity in a room, as capture by the all-purpose sensor.
Credit Liz Reid / 90.5 WESA

“I think what's inevitable is a proliferation of low cost, ubiquitous sensors that computers in the cloud -- or on the internet somewhere -- analyze and provide value from,” Ashton said.

And he said he’s not at all surprised that the pioneering research to develop such a sensor is happening at CMU.

“When we talk about the future of the internet of things, Pittsburgh has a big part to play in it,” Ashton said.

Laput’s sensor isn’t there yet.

A computer still has to be taught what the blender sounds like and what the heat from a stove feels like. That’s a lot of effort for some people, and it requires some level of technical expertise to connect the data from the sensor to a computer that can glean meaning from it.

But making a consumer-ready sensor is not the immediate point of Laput’s team’s research. He said he wants show that this kind of sensing can be done.

“And if you need to be an engineer at Uber, for example, to make it work, then it's obviously going to see limited adoption,” Ashton said.