ClimaSky Innovations

IMAGE-based environmental (sky) analysis with AI.

Comment

Inspiration

Air quality is a pressing issue that affects populations worldwide. Nearly 7 million lives are prematurely cut short every year due to air pollution, making it one of the most pressing threats to worldwide health. While air pollution causes hazardous conditions worldwide, it affects every person living on the Earth in one way or another. Our inspiration derives from the danger and risks that air pollution poses, with the intention of bringing greater awareness to air pollution. This mainly aligns with goal 13 of Climate action in creating a greener future but also intersects goal 3 and 10 of good health and reduced inequalities.

What it does

This program utilizes images to analyze air quality in many areas, especially in areas that are not able to afford expensive air quality monitoring stations. This is an AI for accessibility or health.

How we built it

The program was built using Python's Tensorflow module to create an AI/ML model trained on images of various levels of air pollution. Images were positively attributed to certain air pollution levels and fed into the model in order to train the model. Then, this model was placed into a UX-friendly interface and placed onto a website with ease of access through a photo upload.

Challenges we ran into

We had to develop through the night. One of the biggest challenges was getting the ML model to successfully learn and discern images, as well as integrating two coding languages to post the ML model on a website.

Accomplishments that we're proud of

We were able to successfully develop an ML model that could correctly identify images between bad and good air quality as a team of high school students, creating an app that anyone in the world can use to save the lives of millions of people.

What we learned

We needed to be thinking about code integration as the ML model was developed. This way, the code would be ready to be integrated and it would be not as difficult to integrate into the website once it came time to do so.

What's next for ClimaSky Innovations

ClimaSky Innovations seeks to continue research and development in the sector of AI usage in air quality monitoring. By providing a low-cost solution to air quality monitoring, populations will be more easily informed about the air quality in their immediate region, potentially saving them from health hazards, as well as informing local governance about potential ways for improvement.

Timeline of our work

ClimaSky Innovations has gone through multiple drafts in crafting a perfect application for our users. From the previous hackathon, we were able to connect the frontend and backend using a JSON post request, flask, and React to connect all the functionalities. Our development process from the previous hackathon included improving our machine learning model by finding larger datasets and improving our algorithms to reach a 98 percent accuracy, creating a clean frontend that promoted our message of a clean world, and connecting the frontend and backend to create instant predictions of the air quality.

Built With

Share this project:

Updates