SmartGlow | Devpost

Inspiration

I noticed that many times lights are left on in classrooms even when there's no one inside, and students often turn on artificial lights even when natural lighting from outside is sufficient. This results in a significant waste of energy. To promote sustainability, we created an intelligent lighting system called SmartGlow that automatically toggles lights based on lighting and occupancy conditions.

What it does

SmartGlow works by monitoring the presence of people in a room using person detection integrated into CCTV cameras. Additionally, natural light sensors are placed near the windows to determine whether there is enough sunlight shining to the classroom.

How we built it

The person detection model is based on the latest YOLOv8 object detection model. We assembled a light sensor using a breadboard circuit, which is connected to an Arduino Uno Board. The Arduino IDE communicates with the Python CCTV monitoring software through pyserial communication to toggle LED lights on and off.

Challenges we ran into

One of the challenges we faced was sending occupancy data from Python to the Arduino IDE. However, we eventually overcame this by using pyserial communication, which provided a straightforward method. Assembling the circuit hardware was also challenging, particularly when finding the right resistor value for the LED lights.

Accomplishments that we're proud of

We are proud to have successfully deployed a prototype of this lighting system using Arduino technology. It provides a clear idea of how such a system could potentially work on a larger scale.

What we learned

Throughout this project, we learned coding in two languages, C++ and Python, and how to connect them using serial communication. Additionally, we gained experience in handling input and output devices with Arduino and utilized AI technologies like the YOLO model.

What's next for SmartGlow

In the future, we plan to test the AI model on infrared footage to enable person-tracking even in pitch-black situations, provided the CCTV camera has infrared capabilities. Moreover, we will work on adjusting light brightness levels according to the detected sunlight levels. Lastly, for larger rooms such as office floors, we aim to implement position analysis to toggle lights only in specific areas where people are present.