A Proper Introduction
While the past few weeks have still involved tracking down a few bugs in the system, I wanted to give a proper overview of what this project is, how it came to be, and what I hope it will become. This post will be a bit of personal history, but will seek to explain how my interest in this field has grown over the years.
My interest in home automation & the “internet of things” began as a freshman in college. I’d been exposed to the concept of an automated home before, through exhibits like “Innoventions” at EPCOT in Disneyworld, or the TV (Disney of course) movie, Smart House. Yet it wasn’t until I came across a viral video, showing a practical implication that made me want to jump into the world of connected home technology. Originally posted in 2012, a UC Berkelely student named Derek Low created a video showing how he completely transformed his campus dorm into a fully automated experience, complete with voice controlled lights, motorized window shades, and of course a fully decked-out “Party Mode”. Derek’s video was so popular that his room was even featured on a few news sites. The original video is linked below
Instantly I was hooked on the idea of making my dorm life similar (and possibly one day identical) to what Derek had done as a student at UC Berkeley. Luckily for me, as a freshman in 2013 the market was just beginning to shift to offer connected products to the consumer market. ( I had also just opened a fresh amazon prime account ). So with small steps, the journey began.
Of course, connected products were new, and therefore expensive. The first generation of Phillips Hue Smart Lights had been around for about a year, and still hung at a steep $200 for 3 light bulbs. So turning to the internet, I found my first “smart” bulb on amazon; a single RGB LED bulb, that came with a remote to change the color, dim the brightness, set a sleep timer, and even sync up to music with an onboard microphone. I was instantly hooked.
Next came a few more gadgets..then a few more.. and more.. From a 16 foot strip of RGB LED’s to turn my dorm into a personal club (which got the attention of more than a few other people), to a mini projector to play movies on the ceiling, I was constantly dreaming of how to go bigger. Once my sophomore year came around, I finally had a simple yet powerful automation system in my dorm room. With 4 wifi outlets, 2 motion sensors, and plenty of party and disco lights, we had automatic room lights, an alarm — synced lamp, a party mode, and the early stages of voice control.
The “Party Mode” of the room can be seen in the early portion of this video:
A year later, as a Junior, I had a full campus apartment and larger room to automate. I greatly expanded what hardware I had available — from the Amazon Echo, to a set of 4 wifi controlled LED bulbs, and even control of my TV with the Logitech Harmony Hub. The new hardware was great, but it had one downside. Each device had its own separate app, and only certain devices worked with each other natively. For instance, the Amazon Echo could control the Belkin Wemo wifi outlets I had, but not the LED bulbs I had just bought.
There had to be a better solution. I spend the first half my Junior spring researching ways to integrate everything — searching among device api’s, github projects, youtube videos, and even emailing developers. After much searching, I finally came across one project in particular: Home Assistant. Home Assistant is an open source, python based automation platform that runs on virtually any operating system, that was started by a hobbyist programmer named Paulus Schoutsen. He claims it originally began as a few scripts to have his Phillips Hue lights turn on with the sun (which wasn’t supported natively at the time). Now, the project has over 3,000 github contributors, and supports more home automation hardware than most other commercially available systems or hubs.
At first using this platform seemed daunting, having never used python, been good at programming, or used the terminal extensively. But because home assistant supported both the Amazon Echo as well as the specific LED bulbs I had, I was determined to teach myself how to install and use the system.
Luckily for me, the platform has the most expansive documentation I had ever come across for an open source project, with the addition of user forums and a community live chat. When first installing the platform (on a raspberry pi, which I also had little experience with), veterans to the software were eager to answer my questions on the live chat. 2 months later, after setting up a working system, I even ended up being credited as a beta tester for a new device plugin, after I debugged my motion sensor hardware with another developer (who lives in London) to get it working with the system. Never did I expect I would have that level of membership in a community made up entirely of hobbyists in a completely virtual space.
This brings me to my current setup. Living in a fuller house of campus, I have over 30 connected lights and appliances, including fully controllable lamps, automated fans and blinds, smart overhead ceiling lights, automated entertainment control, LED mood backlighting, temperature, humidity, and smoke detection, and just for fun, and even a color changing toilet bowl light. Every piece of hardware reports its status and can be controlled via the Home Assistant Frontend, which is a web-based dashboard.
Below is a photo of my current dashboard
With additional software running alongside home assistant on the raspberry pi, I’ve been able to control everything via the Amazon Echo and Apple’s Homekit (which includes siri), and expand the system further via integration with IFTTT (If This Then That, A site that connects internet connected hardware with other cloud services).
Despite having expanded the system incredibly fast since its infancy, there’s still many more (almost unlimited) ways I can program the hardware and software to work together to make things more efficient and fun. My thesis project has been 1. Solidifying the installation, integration, and organization of each hardware and software piece into the system, and 2. Looking ahead to future advanced features using this system.
One such advanced feature is condition-based prediction of light, entertainment, and even smell combination preferences. Such a feature would allow a user to return home, and have the home present the user with their favorite combination of light color, music, and scents, based on input data such as daily steps, amount of events per day, or sleep patterns. I believe this physical automated reaction to a user might be able to have a significant impact on a person’s mood, and this is something I’m researching and experimenting with.
Continue following my updates for more progress updates, as well as future video of my setup!
