PillowMate

Inspiration

62% of adults around the world have trouble sleeping according to the Philips Global Sleep Survey. That’s 4.9 billion people.

In Canada, 1 in 2 have sleeping troubles, 1 in 5 don’t find their sleep refreshing and 1 in 3 have trouble staying awake according to the Public Health Agency of Canada.

With so much data being gathered during our waking hours, you would think we would do the same to improve our sleep, a crucial activity we do for on average 8 hours a day. That’s 56 hours a week, 10 days a month, 120 days a year.

Yet, only 8% do according to Statista. So let’s change that!

What it does

Welcome to The Future of Sleep. 💤

Introducing PillowMate, the world’s first smart pillow with built-in 360 surround sound to listen to your favorite music and podcasts, a white noise generator, a smart alarm that gently wakes you up based on your sleep cycle, and monitoring of your sleep, body temperature, heart rate, breathing, snoring, and so much more.

Now in terms of our market size, everybody sleeps. We have 7.9 billion potential customers around the world. Suppose we only service the US & Canadian market, that’s still 370 million customers. And suppose we charge $50 per pillow, that’s a serviceable market size of $18.5 billion dollars, while our total addressable market stands at $395 billion dollars.

Suppose we only capture 10% of the North American market, that’s a serviceable obtainable market of $185 million dollars.

How we built it

• We used a temperature sensor to detect body temperature, a reed sensor along with a magnet embedded into the pillow to detect tossing and turning, and a microphone to detect snoring.
• These were connected to a raspberry pi
• A music player served as an actuator along with a display for the data collected.

Challenges we ran into

• The libraries for the FPGA board we were originally using contained some errors and lacked documentation. We also adapted by switching to a Rasberry Pi.
• Sounds played by the Arduino speakers played distorted sounds, even after two hours of debugging. We adapted by playing from a speaker connected to the Rasberry Pi.

Accomplishments that we're proud of

• We are most proud of having built a hardware design from end-to-end despite having faced challenges in debugging nearly all the equipment we used.

What we learned

• Are the wires connected? Is the file format on the SD card correct? Does this sensor work? Is there a voltage difference between these points? When working with hardware, debugging is much more difficult as you not only have to ensure your code is working, but also that it meets the specifications required by your hardware and that the hardware itself is attached properly.
• Moreover, with hardware it takes much longer to run a program for debugging, both due to the upload time and the requirement of physical input or output.
• Therefore, it is even more important to write careful code as debugging tends to consume even more time.

What's next for PillowMate

• In the future, we would like to introduce an accelerometer or fuse several reed sensor for more precise detection of turning.
• We would also like to learn how different factors such as temperature should affect white noise being played, so that the sensors have a greater influence on the music we play.
• Add 3D 360 surround sound
• Build a phone app that syncs automatically with the pillow
• Get it in the homes of all UofT students so the student body can collectively sleep better!

Built With

• arduino
• chart.js
• python