Lucas Fernando
- @lucasfernando
- | He/him/his
Helping building a better and more sustainable world through tech.
I built a low-cost AI system that can predict air quality from a single photo using the UNIHIKER K10 - a tiny microcontroller built upon an ESP32-S3 chip. By training a lightweight CNN on a public dataset of 12,000 images and deploying it to the board, I created a handheld device that takes a picture, analyzes it, and instantly classifies the air quality index. This project shows how AI on the edge can make environmental monitoring affordable, scalable, and accessible to anyone (as you can see on the images below).
I wrote a complete article about this project on Hackster. Feel free to read it and try to replicate as well. I hope this help you with your next edge computing projects.
Add the first post in this thread.