The new company, Brain of Things, has a pretty simple question at the core of its mission: Why outfit your home with Internet of Things-style devices, when you can simply live inside a robot? The California startup has rigged three houses with a group of sensors and cameras to connect more than 50 home devices into a single responsive network of appliances. That network accrues a stunning amount of data on the owner, and then uses that data to automatically meet the owner’s every desire. The home is pre-programmed and ready for use as soon as the new owner moves in.
“We provide an experience for the owner so they don’t have to do any programming,” Ashutosh Saxena, founder of Brain of Things, told Inverse, adding that the first week can be a little uneasy until a solid foundation of data is collected. “My inspiration comes a large part from self-driving cars and autonomous robots,” Saxena said. “We spend five percent of our lives in a car, and there’s a huge impact, but we also spend two thirds of our lives in the home. But we can’t automate the home because the system doesn’t exist.”
Sure, there are already companies who allow you to turn off the light and adjust the thermostat remotely (take Nest for example, a company whose former VP of Tech, Yoky Matsuoka, is an advisor at Brain of Things). But those disparate smart devices don’t necessarily cooperate with each other. Brain of Things, on the other hand, promises to make everything automatic and connected.
Real estate developers have taken note and expressed interest. All that connectivity can justify around $125 in additional monthly rent while costing the property owner around $30 every month for upkeep. That’s a decent value proposition, especially for reluctant landlords.
For Saxena, programming is the old way of doing things. Machine learning, or deep learning, is the way of the future. The house learns when the owner wants the lights on, the curtains open, and the room at a cool 65 degrees. It also knows who is at the door, and if it’s the mailman, the house knows if you want that package or not. And, of course, it’s connected to the owner’s phone, so people can do things like compulsively watch their dog eat at random points in the day.
And this isn’t just for Silicon Valley geniuses and the tech savvy side of the population. In fact, it’s designed for the average person: “That is where the real market is,” Saxena said.
Average people might be a harder sell than the uber techy, though. Especially an average person plugged into media. Take Disney’s 1999 movie Smart House, in which a computerized, home not unlike a Brain of Things house, goes crazy and becomes an overbearing control freak. It’s a movie Saxena admits he is all too familiar with.
“It’s a very practical concern for any A.I.,” Saxena said. “It has to be constrained in some way, so what we have done is make very, very strong constraints on where the system can learn and what the system can learn.”
There’s a reason why even people who closely follow Internet of Things’ progress didn’t have Brain of Things on their radar. Saxena and co kept tight lipped about their project before this week. They wanted to go public with their ideas when they had a ready product, because as Saxena says, “You can’t sell a car with only two wheels.”