Over the past decade or so, as AI has become more common in the background, it got a lot more common for consumers with smart homes to build smart security systems on their own – rather than having to pay high fees to established old-school security companies.
Given the fact that home security has become so much more democratized, you can bet that AI features are coming to those systems too. In fact, there are plenty of AI-based features that have already started rolling out to home security devices, ensuring that home security systems are more advanced and more helpful than ever. Here’s a look at some of the most common of those features – and how you can take advantage of them.
Object and person recognition
Object recognition is getting a whole lot more advanced. Over the past few years, security companies have been building better facial recognition into their cameras, ensuring that you always know who's at the door near your home. Those systems are going beyond that too, able to recognize objects like packages on your doorstep, vehicles parked in your driveway, and even license plates from suspect cars.
These features have tangible benefits; not only will you be able to know who's at your home and, but you'll be able to eliminate false alerts due to the movements of trusted individuals, ensuring you only get notifications when the system detects someone it doesn’t recognize. No longer are systems constantly sending notifications just because they detect a tree moving in the breeze. Object recognition ensures that they can filter out movement that's irrelevant to home security.
The majority of smart security cameras offer object or person recognition of some kind, especially when it comes to common objects like packages and people that a camera sees regularly.
Video analytics and summaries
Analytics from security cameras can go beyond simply detecting objects and people. Video analytics can allow cameras to better recognize when someone might be behaving suspiciously around your home. For example, cameras could identify that someone is carrying a weapon or something they might use to break into your home, then notify you of that behavior. Generative AI will allow those notifications to be richer and more helpful as well, thanks to the fact that it can help systems create summaries of activity that's detected with specifics about what the cameras saw.
Google has started leveraging AI for these analytics and summaries in the Google Home platform, and it's likely that others will follow suit, with Google more widely rolling out the feature to more cameras and systems.
Advanced automations
Smart home automations can be incredibly powerful as they allow you to set up your home to work without you needing to actually do anything. AI will help make those automations easier to set up and run more efficiently, largely through natural language recognition, which will allow you to set up automations with a simple command. For example, you could tell your smart home system to "always lock the door five minutes after I leave." This could be done without you having to scroll through menus, set up individual devices, and so on.
A feature like this is already rolling out as part of Google Home with certain Nest devices. It runs on the Help Me feature in Google Home, which allows you to text describing the automation you want to set up. The feature is expected to widely roll out by the end of 2024.
Better voice control
Smart security systems are often controlled with voice commands, allowing you to arm a system or camera using a service like Google Assistant or Amazon Alexa. Google has announced that it's supercharging the Google Home platform with AI features based on Gemini, which will allow you to find out information about your smart home using the devices that you already have set up. The example that Google gives is that you'll be able to ask Gemini if the kids left their bike in the driveway, after which Gemini will be able to use a security camera that you might have set up in the driveway to find the answer to your question. This is another feature that Google has shown off and is expected to roll out widely by the end of 2024.
