A few months back (over a year ago) a student and I decided we were going to make a “magic mirror”. They have been all the rage lately on Reddit, GitHub, and a few other techie websites. I digress, what I am trying to tell you is we started and stopped more time than I could count. And after countless talks about what platform we were going to build this thing on, either Android or Raspbian (Raspberry Pi OS) we have finally deliberated and the verdict was Raspbian. F I N A L L Y.
So, now that we had our platform we needed to build a test machine. So we jumped on Amazon and we ordered a Raspberry Pi 3 kit. We ordered the three because we wanted built-in WiFi so that we could decrease the amount of clutter behind the screen. We also thought it was convent to buy the Raspberry Pi 7-inch touchscreen monitor. This would make it easy to test right from our laps. Let’s face it we are tech people and we are lazy, this screen allows us to trouble shoot from the couch in my office.
We also just purchased this enclosure for working on. It is a simple easy to use enclosure/stand that will allow us to add components on if needed since it is wood. All we have to do is glue and screw.
With the hardware out-of-the-way we now had to settle on how we were going to exactly make this mirror work. Our thought process was a little jumbled but in the end we narrowed the software requirements to the following;
- Voice Recognition
- Touchscreen (possibly)
- Motion activated (still in flux)
Before I get into the break down of the qualities our mirror or as we call it project glass would soon take on, I think I should take a moment to explain or end goal for this build.
Our hope is to build a finished product that is both sleek and good-looking but ultimately looks like a mirror. We then hope to install it in the front lobby of our school where Admissions has their guests waiting for their interviews and tours.
“Wouldn’t it be cool if when our guests could ask the mirror on the wall something specific and it would return and answer for the inquiry?”
Trick question, of course it would be C O O L !
Simple, questions like; “What year was this school founded?” or “How many day-students attend Suffield Academy?”
Any how this is our end goal. The next few weeks will be challenging and I am sure we are both going to learn a lot, I believe the result will be totally worth it.
Alright, onto how we plan to achieve this. We are going to attempt voice Activation and recognition with Amazon Echo’s voice API, which is open source. You can sign up here. They even have a Dev kit available to buy. This kit comes with Conexant AudioSmart™ 2-Mic Development Kit. This kit allows you to quickly create a working prototype due to these handy features:
• Dual Microphone Hands-Free Voice Interaction with Smart Source Pickup™
• Low Power Embedded Alexa Wake Word Engine
• Full Duplex Stereo Acoustic Echo Cancellation (AEC)
But, due to the fact that we are trying to complete this build for sub $200 this option is off the table. Who knows, maybe if this project goes well we will get lucky someone will make a little donation (wink wink world).
Touchscreen, sounds fancy right?
Don’t worry it is. Just kidding, I am still trying to figure out how this would work. Do I connect a small touchscreen beneath the mirror for a fail safe, or to act just as an onboard keyboard and possibly add web browsing to Project Glass?
This one is still up in the air. But here are some pics of what the RPi 7-inch touch screen looks like. Along with some easy steps for setting it up.
Okay I am going sum up the last few bullet points in a few short sentences. Ready? Okay, here it goes…
I am going to make my partner (the student who is actually building this figure out all the specs).
In all seriousness, this is a student led project. I am just here for moral support and helping if they get stuck, and of course documenting our steps.
Project Glass will officially begin this Friday. I will of course keep you all updated along the way.