Back to Blog

How AI Is Powering the Next Generation of Smart Homes

For years, consumers have been flirting with “smarts” in the home. Only recently has the confluence of better, cheaper and smaller sensors, low-cost cloud computing, and ubiquitous, persistent broadband connections made “Smart Homes” a reality. We can see builders and homeowners taking advantage by installing connected TVs, thermostats, light bulbs, door locks, and other objects that increase safety, comfort, and convenience.

The next evolution is underway as products are able to gather massive amounts of data to train artificial intelligence and machine learning algorithms to provide personalized experiences for homeowners. Sometimes the AI can be seen overtly; sometimes it is invisible in the devices themselves.

Remember the Roomba vacuum cleaner from 2002? While functional in the cleaning department, the Roomba would bump into walls and furniture because it could not map the area needing to be cleaned. Now we have the Dyson Eye 360 using LIDAR, the same machine learning and sensor technology in self-driving cars, to map rooms, and plan cleaning routes accurately.

Amazon has released an Alexa-powered microwave. The cooking process is straightforward: put a potato in the microwave and tell Alexa “one potato” and the microwave starts immediately. There are even smarter appliances like the June Oven. Besides being a convection oven, air fryer, slow cooker, broiler, and toaster all in one, the June Oven uses built-in cameras with computer vision and image recognition to identify the item, and recommend best cooking methods and time. Over time, the machine learning algorithms learn how eaters like their food cooked and will make more accurate suggestions.

App and Bluetooth-powered keyless door locks have been around for a few years. Companies like ZKTeco and Corum are integrating cameras for facial recognition capabilities to allow homeowners to unlock their doors with nothing but a smile. The locks can recognize faces within 0.2 seconds while providing backup mechanisms to open doors, such as key fobs, codes, and, of course, real keys.

With all the connected devices and data flying around homes, there is a heightened need for security and privacy. Many smart home device manufacturers do not implement best practices of security by design when writing software, often leaving baby monitors, routers, and gateways to be hacked. There are even search engines to find vulnerable devices by IP. To prevent unwanted data leakage, the Bullguard Dojo is a device that creates a smart firewall, funneling and filtering data from all connected devices. Using machine learning for behavioral analysis, network traffic analysis, and threat detection, and combined with user input, Dojo decides what data can leave the house.

For home builders, selling a pre-packaged system of intelligent devices presents challenges. For example, while Amazon is making inroads working with builders to get its set of products in new construction homes, Alexa may not be the best option for homeowners. The three players in home automation — Amazon, Google, Apple — all have their own protocols. Devices may support one or any combination of protocols. Similar devices may serve different purposes. The Amazon Echo, Google Home Hub, and Apple HomePod may all be voice assistants, but the HomePod is positioned more as a speaker for music. Homeowners may have preferences concerning integration with their mobile devices, such as iOS vs. Android. Finding the right combination of devices is as unique as homes themselves.