Thanks to the Internet of Things, the home of the future could be full of smart fridges, lights, thermostats, TVs, toasters, doorbells, picture frames, and even whiskey bottles. A team at Carnegie Mellon University is trying to untangle that web, by developing a single device that uses a suite sensors to monitor an entire room's worth of regular appliances.
The prototype, which the team calls a Synthetic Sensor, plugs into a regular wall outlet, and keeps watch over the room by way of nine sensors that commonly crop up in smart home devices. It detects sound, light, vibrations, heat, temperature, and electromagnetic signals, then uses machine learning algorithms to figure out what that data means.
"The idea is you can plug this in and immediately turn a room into a smart environment," says Gierad Laput, co-author of the study. "You don't have to go out and buy expensive smart appliances, which probably can't talk to each other anyway, or attach sensors to everything you want to monitor, which can be both hard to maintain and ugly. You just plug it in to an outlet."
The algorithms can use that data to identify when and how certain devices are being used, and they're powerful enough to tell the difference between a coffee grinder and a blender based on the sounds and vibrations they make. More subtle changes won't escape notice either, with one demonstration showing the system detecting a user writing on a whiteboard, then erasing it.
While it can't control the devices it's monitoring, through context clues, the Synthetic Sensor can tell where in a cycle a device might be, listening out for when a microwave is running, interrupted or finished, or pinging your phone if the oven's been left on. Like other whole-house monitoring systems, it could know when the occupants are sleeping, up and about, or away at work. With some number-crunching, it could calculate how much water is being used by how long the tap has been running, or how much toilet paper is left before you need to restock.
"It can not only tell you if a towel dispenser is working, but it can also keep track of how many towels have been dispensed and even order a replacement roll when necessary," says Laput.
Wherever this much monitoring is done, concerns about privacy won't be far behind, and interestingly, the group deliberately hasn't put a camera into the device for that exact reason. Since it could be gathering some pretty sensitive data, the device processes and stores it all locally, to reduce the chances of it falling into someone else's hands.
Of course it's just a prototype at this stage, so there's no word on when, if ever, the Synthetic Sensor could make it to market. The research was published online, and the team is presenting the device at the Computer-Human Interaction conference in Denver this week.
The system can be seen in action in the video below.
Source: Gierad Laput
