Machine Learning Baby Monitor, Part 2: Learning Sleep Patterns

The first lesson a new parent learns is that the second you think you’ve finally figured out your kid’s patterns — sleeping, eating, pooping, crying endlessly in the middle of the night for no apparent reason, whatever — the kid will change it. It’s the Uncertainty Principle of kids — the mere act of observing the pattern changes it, and you’re back at square one.

As immutable as this rule seems, [Caleb Olson] is convinced he can work around it with this over-engineered sleep pattern tracker. You may recall [Caleb]’s earlier attempts to automate certain aspects of parenthood, like this machine learning system to predict when baby is hungry; and yes, he’s also strangely obsessed with automating his dog’s bathroom habits. All that preliminary work put [Caleb] in a good position to analyze his son’s sleep patterns, which he did with the feed from their baby monitor camera and Google’s MediaPipe library.

This lets him look for how much the baby’s eyes are open, calculate with a wakefulness probability, and record the time he wakes up. This worked great right up until the wave function collapsed the baby suddenly started sleeping on his side, requiring the addition of a general motion detection function to compensate for the missing eyeball data. Check out the video below for more details, although the less said about the screaming, demon-possessed owl, the better.

The data [Caleb] has collected has helped him and his wife understand the little fellow’s sleep needs and fine-tune his cycles. There’s a web app, of course, and a really nice graphical representation of total time asleep and awake. No word on naps not taken in view of the camera, though — naps in the car are an absolute godsend for many parents. We suppose that could be curated manually, but wouldn’t doubt it if [Caleb] had a plan to cover that too.

12 thoughts on “Machine Learning Baby Monitor, Part 2: Learning Sleep Patterns

  1. > “This worked great right up until the baby suddenly started sleeping on his side”

    That’s the tough part about learning anything about your child in the first year: in two weeks it’ll likely change.

  2. something we found with our granddaughter. There’s something called Signing Time ….. it’s designed to teach babies sign language & believe it or not it worked… the little one could actually tell us simple (at first) things like more of something, wet, tired, etc..
    Of course it requires the parent(s) to watch as well or it’s really going to piss of the kid.. :)

    This is definity a cool & adaptable project… Thanks…

    1. I will place a warning on teaching babies sign language. If your child can clearly communicate with you via sign language one of the strongest motivators for language development gets thrown out the window. Of course your baby can learn sign language the damn things are smart as hell but you need to help them prioritize.

      Personally if you want to watch a fun development, watch newborns learn how to see. Slowly the develop an understanding how to blink, how to focus, how to keep their eyes tracking the same direction. It all happens in just the first couple weeks of life but it was incredible to watch.

      1. it is….. luckly I was able to do this part time.. :)
        at least with Eva, it didn’t interfere with her language skills, she was right “on time” I’m told…
        your mileage may differ… :)

      2. I don’t think loss of motivation from being able to communicate basic needs is a thing. Babies have a pretty insatiable drive to learn.

        I looked for literature on this, and there seems to only a few papers hat actually gathered data with a reasonable sample size.

        Goodwyn et al, 2000: 100ish babies between the control and sign-taught groups. They claimed some pretty solid results in favor of signing, but there has been criticism (Paling, 2007) that they didn’t do their due diligence in eliminating experimental biases.

        Kirk et al, 2013: 40 kids. No real difference in spoken language between control and signing groups.

        Seal et al, 2014: 16 kids (man, these sample sizes are tiny). Signing group saw more babies reach word milestones, but not statistically significant (not surprising with only 8 kids each in signing and control groups)

        There are tons of meta-analyses that realize that the actual studies done so far have small sample sizes, and there are a bunch of shoddy papers with unsupported, yet far-reaching, conclusions with just a few kids. Like almost all literature on raising babies, good data is rare, and recommendations are made vociferously (just see the AAP’s most recent, extremely strong recommendations and how flimsy of a basis they’re built on). The state of science in this field embarrasses me every time I look at it.

        Regardless, even with the relative lack of solid data on the effects of baby sign language on oral language learning, what little data there is shows absolutely no signs of a negative impact, and that there maaaaaaaaaay be positive impact.

        When you look at its other benefits, one’s baby being able to communicate needs long before they’d be able to via speech, I don’t think it’s responsible to advise against teaching it due to unsupported theories of learning impairment.

          1. Make that a sample set of three! Our two kids have thrived with it, there was much less frustration after we introduced sign language to the first one. The second one we’ve introduced it right away. Their spoken language did not suffer at all. One might even argue that their contextual awareness only grows, increasing the desire to talk! Now they can interact with adults, and get rewarded for communicating, how exciting is that! It was a while ago, they’re 7 and 9 now and fully bilingual.
            Oh, side effect: they could tell us when they needed to go to the toilet at only a few months old! No poo diapers anymore!

  3. “this machine learning system to predict when baby is hungry; and yes, he’s also strangely obsessed with automating his dog’s bathroom habits.” ….. you wouldn’t want to get these projects mixed up :(

  4. hmmm good to see honestly. For me it was just a hunch happy to be shown wrong. I was worried because I knew a baby who only signed until ~2.5 years and I felt like his ability to sign more and more words just kept him demotivated but that could be a million other things to be fair.

    Babies are wild learning machines.

    1. I agree that babies are wild learning machines, but let’s not forget that every adult is a wild TEACHING machine. Let’s say you wamted to learn Mandarin, or Hungarian. Now imagine that literally everyone you met smiled at you, and talked to you in that language. Then repeated words, pointed at things and named them. Gently corrected your pronunciation. That’s what it’s like to be a baby and I reckon people in their 40s would learn a language pretty quickly too if everyone around them made the same effort that they do around toddlers.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.