HomeAI: The Future of the Smart Home

HomeAI was born from two goals:

  • Smart device integration with an AI which learns about, and can predict, the user's interactions with their home
  • A more efficient usage of energy due to the ever evolving AI.

These goals stem from our passion for a more comfortable experience with our own homes while also being as "green" as possible.

We wanted to make a home which was truly automated. By "truly automated" we mean that we wanted no manual inputs/outputs from the user. Our AI is uses its prediction algorithms to anticipate certain events and it is these that will be the input and outputs of your smart home.

What it does

HomeAI uses a predictive AI to take in data from various inputs. Your smartwatch, Google Nest, Philips Hue lights, and other devices all work in tandem to feed data to our AI. In the end, it aims to adjust light, temperature, and door locks to be in the optimal setting based upon your personal preferences, time of day, and status of the home's occupants.

Additionally it will passively conserve energy whenever possible. By understanding how you use your home, it will make sure that there is no unnecessary waste in energy. To understand this a little further and to see quantifiable savings we put our system to the test. We found out that this technology will control the use roughly 212.5 watts (including your homes HVAC system) of your day to day power usage. (versus the average usage of 840 watts by a home in New York state*) This puts the user in control of around $7.67 a week* (or $30.68 a month) of possible savings. Within two years of setting up your smart home the energy savings will have surpassed the initial set up cost of the smart devices.

How we built it

HomeAI was built using a plethora of technologies, some of which include Amazon Web Services resources and many smart home devices. The programming languages include: java, C++, Javascript, html5, css3. We started by understanding how to control all of the smart devices under one platform. Once this was accomplished, we shuttled the data over a websocket to be captured by our AI and allowed it to process and analyze the data. This data was eventually stored in Dynamo DB which can be pulled by our user at any time though our secure, web portal.

Challenges we ran into

HomeAI ran into a few challenges such as learning the various Amazon Web Services which it required to run properly and how to send the data back to the AI.

Accomplishments that we're proud of

We're proud of the whole project; it was a unique challenge but overall one we're passionate about.

What we learned

We learned a lot about the resources available from Amazon Web Services, especially their database Dynamo DB.

What's next for HomeAI

HomeAI's future is currently unknown but potentially we will expand upon the idea to work even better than it does currently.


*Data points from the Department of Energy (2013); cost scaled to account for inflation.

Built With

Share this project:
×

Updates