Smarter Homes With Artificial Intelligence

Machine Learning Will Make Devices Much Smarter

 

Much has been written and will continue to be written about artificial intelligence (AI).  While the term artificial intelligence was coined in the 1950s, the more recent term that is talked about almost as much is machine learning, which is what is driving the rapid advances in AI.  Machine learning is the ability of computer algorithms using immense computing power to recognize voices, objects, speech, and countless types of patterns as well as making relationships between pieces of information that previously had to be explicitly programmed by a human.

If that sounded somewhat complicated, it is.  The mathematics behind the algorithms are complex.  And the algorithms are always being tuned to be “smarter,” in much the same way humans can learn from mistakes. Machine learning is the fundamental technology that will give us fully autonomous cars in the next five years, to name but one large area it will impact.

But AI is also coming to home automation and smart devices that we use every day.  Read on to learn about some artificial intelligence applications that are available right now for your New York City residence.

SEE ALSO: HAL Is Here – Almost

 

Voice Assistants

Last month we discussed smart voice assistants like Amazon’s Alexa and how they are rapidly being integrated into home automation.  Machine learning is what enables services like Alexa, Google Assistant, and Apple Siri to understand your queries and turn them into something actionable.

In the past, voice control systems – like those in cars – were rigid; they needed specific, clearly enunciated commands in a defined order to understand you.  With machine learning, voice assistants now understand natural language constructs – although within some limits.  The powerful part of machine learning is that the more Alexa, Siri, and other voice assistants are used, the smarter they get.  The more different types of queries they field, and accents and languages, the better they get at understanding what you want.  So while Alexa may not completely get “Alexa turn off my kitchen island lights in 5 minutes” the first time, rest assured that the more people that use it for those purposes, the more it will learn what you want to do as it recognizes your speech and what you do.

Machine learning is a voracious learner. Vast amounts of data are needed to make machines learn effectively.  So keep on talking to your Alexa, Google, or Siri; you’re part of their education.

 

Smarter Video Surveillance

One of the most exciting advances to make smarter cameras is in applying artificial intelligence to recognize objects and people in pictures.  If you happen to use Google Photos to back up your photos from your smartphone and computers, there is at least one reason why it’s free.  All the billions of photos that Google stores helps its AI algorithms learn how to detect things in the pictures.  That’s one reason why you can search your pictures by typing “dog” or “Eiffel Tower” and instantly find all the related ones. And that image recognition technology – tuned by all that data we provide – is invaluable for use in many other applications.

The same type of technology is being applied to make smarter video security cameras.  If you have one at home, and it’s set to notify you when there’s movement, you may get many notifications per day when your dog walks by.  The latest models can be trained, so they can recognize that the object is a dog rather than a burglar.  Some surveillance camera vendors are now employing cloud computing power to identify objects to locate events in a continuous video feed quickly.  So if the suspicious truck in front of the building says FedEx, the software can deduce it’s most likely a delivery.

Smarter Home Automation

Ultimately the same technology will make for even smarter home automation.  For example, today Google’s Nest thermostats can monitor your location through your smartphone and adjust the climate control accordingly to save energy.  It can also learn your patterns of coming and going.

In the future, many more smart devices will have access to the same patterns.  A water heater might not heat until it estimates when it needs to because there will be demand for hot water.  A dishwasher might detect it’s full (with an internal sensor), know that you’re at work, and turn itself on to be sure the contents are clean when you arrive home.  The possibilities for automation are endless with increased intelligence, making a home truly smart.

 

While we wait for the future, let us show you what you can do with smart home automation today.  Reach out to our experts today.