You may have heard the term “machine learning.” Many businesses are pretty excited about the concept.
However, to the average consumer, machine learning can be difficult concept to wrap one’s mind around. All the talk of massive data gathering operations, challenges with big data, complex algorithms reminiscent of artificial intelligence, and business intelligence upgrades ring a bit hollow. At the same time, the average consumer stands to gain quite a bit from the advances being made in machine learning technology. So what’s the best way to demonstrate the benefits of machine learning to someone who may not be well-versed in the latest technological trends? By showing how machine learning applications are shaping the next generation of homes.
The trend of home automation has been around for a few years now. Much of the progress made in that area, though, has been through manual programming or remote control. Many home automations need to be set up by the homeowner explicitly telling the system when to raise the temperature on the thermostat, for example. The user can also control these systems with their smartphone, like turning on the lights minutes before pulling up in the driveway. These systems are certainly a more convenient way to interact with a home, but they only scratch the surface of home automation potential. It’s with machine learning where that potential will be reached, and much of it will be done with little effort on the part of the user.
Part of the increasing emphasis on machine learning in home automation is the recent rise of the Internet of Things (IoT) along with storage technologies like a flash storage array. Devices connected to the IoT can gather all sorts of data on people and send that information to other devices and databases. It is this collection of information that can help drive machine learning applications in the home. Machine learning algorithms, as can be seen in the name, learn over time as they constantly update their processes to reach a desired result. In the case of machine learning in home automation, devices can learn the behavior of residents and respond accordingly. For example, an automated home can learn one resident opens their bedroom blinds when they wake up every morning. Based off of that, the home will automatically open the blinds at the exact time they get up.
The potential applications of this are nearly limitless. We’re already seeing the first wave of such home automation. One startup called Brain of Things is constructing apartments it’s dubbing “Robot Homes”. These living spaces contain numerous motion sensors that can track where people are in the dwelling and what they’re doing. Based off of this information, the Robot Homes can automate many of the apartment’s systems like lights and appliances. In a sense, this use of machine learning applications can anticipate the resident’s needs before they even arise. Consider it a more responsive smart home that can also send notifications if anything is detected that’s out of the ordinary.
But that’s just one example of the next generation of home automation. Researchers at Stanford and Cornell recently unveiled what they are calling the Watch-Bot, a robot that comes with machine learning algorithms and sensors designed to observe people in their living quarters. Based off of data collected from videos found on sites like YouTube, the Watch-Bot can determine if you leave objects where they shouldn’t be or forget to close the fridge door. It may sound more like a nanny than a robot, so it remains to be seen how receptive consumers would be to such a product, but the fact that machine learning allows it to learn when things are out of place shows how far the technology can go.
Home automation involving machine learning can also be used in security systems, smart plumbing, heating and cooling systems, and so much more. The many machine learning applications that can be applied to manufacturing or financial systems may fly over the heads of most consumers, but seeing it work inside where we live brings it all much closer to home. Once consumers experience the convenience for themselves, they’ll be more willing to let more machine learning into their lives.