Inspiration

Ever wanted to have a butler AND a watchman but didn't have the money to afford either? Well, we did and so we built Holmes, your very own intelligent home assistant.

What it does

  1. Captures photos of all incoming people at the door of the house and verifies their identity against a pre-populated database of recognized individuals using the Microsoft Cognitive Services API. If verified, the name is added to the Initial State database with the following keys: Name, Status (inside or outside), Recognized (yes or no), Time of entry, Number of people in the house. It also has the ability to add the recognized image to the database to further train itself and become 'smarter'. If the individual at the door is not recognized, a text message with an image of the unidentified individual is sent to the owner via Twilio.

  2. Tracks the total number of people inside the house at any given point of time by tracking both incoming and outgoing traffic.

  3. Once a person is granted entry into the house, he/she is greeted (by name) by Alexa and asked for choice of music to play.

  4. The owner can monitor all said data in real time on his/her Initial State dashboard.

The entire program runs on Intel-Edison i.e. the Edison works as the main computer for the execution.

How we built it

Used Python as our main language and installed the latest version on Edison for compatibility. The program runs entirely on Edison and works remotely too.

Challenges we ran into

  1. Finding a room to begin our hack.
  2. Compatibility issues with the Edison.
  3. Connecting the camera to Edison and making it capture images in a fixed interval of time.
  4. Producing text to speech output on Alexa

Accomplishments that we're proud of

Since it was our first time hacking on hardware, we were quite excited about how we would go about it. We are especially proud of how we managed to integrate 3 hardware devices (Intel-Edsion, Camera, and Amazon Echoand 4 APIs (Microsoft Cognitive Services, Bing Text to Speech, Initial State, and Twilio).

What we learned

  1. Integrating hardware with our programs and scripts.
  2. Converting text to speech and producing the output on Alexa.

Built With

Share this project:
×

Updates