Devices like Nanit’s artificially intelligent baby monitor are designed to outsource much of that brainwork. It’s a $349 night-vision camera that hangs over a cot, using computer vision and deep learning to monitor your little one’s sleep. The little gadget is relatively small, but it packs a microphone, speaker, camera, nightlight and temperature sensor inside its body.
The company describes itself as the “Tesla of baby monitors,” offering artificial intelligence and machine vision. In addition, the device is designed to resemble a particularly elegant-looking lamp and hide its geekiness in plain sight. The company’s pitch also encompasses some of the best security for a baby camera available on the market, should you be worried about hackers.
Nanit remains in position with the help of a five-ish-foot-tall floor stand that needs to be screwed into the wall behind with a neck brace. As with a self-assembly IKEA lamp, you screw in a pair of angled feet to help the pole balance. Word to the wise, however: The feet fixings aren’t the best, and their screw ends began to thread as soon as I began attaching them.
Due to federal regulations governing the position of cables near cots, Nanit’s USB cable is buried in plastic piping. These segments are 20 inches long and are designed to look unobtrusive when placed against your baseboard. Unfortunately, if your outlets are relatively close to the cot, you’re left with an unsightly pile of tubes that you’ll need to hide behind furniture. The company has, however, promised to work on a more elegant solution to this problem in the future.
Using your smartphone as a display, Nanit will work as a regular ol’ video baby monitor, piping sound and vision from the cot 24/7. In addition, a built-in night-light can be activated and deactivated from inside the app. Users can also see temperature and humidity readings for their nursery, as well as save snaps of the sleeping kid to the camera roll.
There is one surprising omission from Nanit’s feature list: the lack of a way to use the device as a walkie-talkie. Plenty of cheaper monitors (including my own) enable you to chat to your kid via the camera’s built-in speaker. On one hand, settling your child remotely almost never works, but it’s useful for asking your partner for their drink order from the other end of the house.
The other, more important half of Nanit’s pitch is that it’ll offer you insights above and beyond a garden-variety baby monitor. With machine learning and computer vision, the device tracks your child’s sleeping pattern and can interpret that data for your later perusal. Each morning, you can find out how much sleep your kid had and how long it took them to fall asleep.
You’ll also get a highlights reel, showing you an edited-down, sped-up video of all the notable incidents that took place during the night. These reels are never more than a minute or so long, but they’ll show you if your kid quietly woke up and fell asleep during the night. The nature of the clip, however, means that you’ll be mentally playing this tune as you watch the video.
The data’s not just for learning how sleep-fucked you are, but also for planning your kid’s day to suit their needs. If they’ve slept for 11 hours without waking, then you can take them to play gym and lay off the midday naps. If their night (and, by extension, yours) was terrible, then you can treat them with kid gloves during the day and put them to bed earlier the next evening.
Another thing that Nanit offers is a notification stream, which lists every documented event that takes place during a night. Should you need to examine a particular incident during a very rough night, you can scroll through every data point the device absorbed. There is the risk, however, that you can drown in so much data that you’ll never get the real benefit of it.
There’s also a general problem with computer vision (rather than Nanit specifically) that provided a couple of amusing errors. While these devices “see,” they don’t necessarily understand, which can cause a pretty big problem. One day during testing, the app’s dashboard (pictured) told me that my kid had spent six and a half hours asleep during the day. Which was interesting, because she’d been out at the play gym all day and had yet to return home. The camera had, unfortunately, registered her sleeping bag — laid out on the bed for that night — as her. Presumably, as the light shifted through the day, the system had thought it saw enough motion to believe she was there.
All Credit Goes To This Website: Source link