FireMonitor.AI

• 

  The one and only Fire Monitor

• 

  FireMonitor.AI

Inspiration

We wanted to use powerful neural networks to observe and interact with the world in a novel way that also saves lives.

What it does

It's your personal AI that detects the exact moment a fire starts in a building, alerts you and residents via text message, checks that no one is left behind, and provides real time updates to first responders through both audio and visual data. Given multiple video streams of rooms, this web app detects if a fire becomes present in any of the videos. Using intelligently trained neural networks through the clarifai API, the image of a fire can be detected and an alert text message is immediately sent to building residents. Additionally, a floor plan map updates in real time displaying what room the fire is in as well as what rooms still contain people, and displays exit routes. The application gives real time updates to first responders, and provides critical information about the environment.

How I built it

Clarifai's API detects, among other things, fire or people in images, and the web app is built with python and flask framework. The building map updates in real time using jquery. Text data is sent to the raspberry pi where it is turned into speech.

Challenges I ran into

Processing video in real time, updating webpage in real time.

Accomplishments that I'm proud of

Building an application that has real usefulness.

What I learned

I learned about combining front end and back end, working with others on a git repository, as well as the value of presentation.

What's next for FireMonitor.AI

Use Clarifai further and train data sets specifically to detect fire even more accurately and quickly. Process live video directly, maybe using Facebook live stream. Talk to fire departments and building safety managers about next steps.

Built With

• bash
• clarifai
• css
• fireofinnovation
• flask
• html
• javascript
• jquery
• python
• raspberrypi
• texttospeech