GreenLight

Landing Page
About Us & Contact
Login Page
Home Page

Inspiration

We were inspired to create this project as our team believes in the judicious and efficient use of resources. We seek to implement our beliefs by using current technology and solid ideation and hence came up with the idea of solving the issue of electricity being wasted by lights being left on in a vacant room.

What it does

Our Project uses a CNN-based model that detects human presence in a room. If it detects that a room is vacant with the lights on, It sends an alert on our website, pinpointing the exact room that needs the lights switched off. The alert sent on our website consists of the room name, the status of the lights, and the camera feed of the room.

How we built it

Our project used the video footage of a CCTV camera in the room to determine the status of the room. We used YOLOv3 pre-trained models for human detection and OpenCV for light detection. After the status of the room is determined, the data is passed through an API. The front end of the application calls this API every minute to determine the updated status of the room and displays it on the screen.

Challenges we ran into

While building the application, the main challenge we faced was how to integrate the CNN model and the backend of the application. There is not much support available for the problem we faced, on the internet, so we had to figure out a lot of solutions on our own, and after failing multiple times, we finally found a solution that works.

Accomplishments that we're proud of

We are proud of the fact that we were able to implement our beliefs and devise a solution where we smoothly integrated the various types of technologies. We also take pride in the fact that we were able to learn so much and complete this project in such a small amount of time. Last but not the least, we were able to take a stride in the direction of saving the environment.

What we learned

Starting with the ML part, we learned the limitations as well as the various features of YOLOv3 and dove deeper into the concepts of AI detection. Learning about API data passing as well as the smooth integration of the API from the backend into our an displaying it accordingly.

What's next for GreenLight

We plan to scale our project to various institutions and workspaces. Furthermore, our API can be used by embedded systems to turn on and turn off the lights automatically.