AIR - Your Personal Asthma Assistant

Sign Up Page
Home Informational Page
Risk Map
Environmental Details of Location
Event Logging

Inspiration

As one of the 25 million Americans living with asthma, our team understands the constant mental burden of wondering if the next location, weather change, or air quality shift might trigger an attack. This inspired us to harness AI technology to create a proactive solution, especially knowing that asthma accounts for 1.6 million emergency room visits annually in the US alone. By combining real-time environmental analysis with personalized machine learning insights, we're not just building an app - we're creating a way for millions of asthma patients to confidently explore life while staying protected by data-driven predictions.

What it does

AIr is an AI-powered asthma companion that:

Tracks & Learns: Allows users to log asthma episodes alongside location information and automatically tracking environmental data (air quality, PM~10~, weather, CO and NO~2~) to identify personal trigger patterns. The ML model gets smarter with each data point, understanding your unique asthma profile. Predicts & Alerts: Analyzes your planned locations against real-time environmental data to send personalized risk alerts. If you're heading to a park during high pollen season or a city with poor air quality, you'll know before you go. Empowers & Prevents: Provides actionable insights and recommendations, helping users make informed decisions about their daily activities, travel plans, and preventive measures. Instead of reacting to asthma attacks, users can proactively manage their condition.

The key differentiator is that the system becomes more personalized over time - it's not just generic asthma advice, but tailored insights based on your specific trigger patterns and historical data.

How we built it

We built AIr using React Native with Tamagui for a smooth, native mobile experience across iOS and Android. The app's backend runs on FastAPI, chosen for its high performance and easy integration with our ML pipeline. Our predictive models are powered by scikit-learn, which processes user-specific asthma data alongside environmental factors from Open-Meteo's API to generate personalized risk assessments. Firebase handles real-time data synchronization and user authentication, ensuring users always have access to their latest predictions and alerts. This stack enables us to deliver real-time, personalized asthma insights while maintaining a responsive and intuitive user experience.

Challenges we ran into

We initially had trouble running Expo since Expo Go was being blocked by the network and we had never completed a project with it.

Accomplishments that we're proud of

We are proud that we were able to turn our idea into a cross-platform mobile app. That is something nobody on our team has done before and it was a great experience and challenge.

What we learned

We learned a lot about asthma data and how personalized it can be. We also learned how to develop with Expo and all the nuances that go along with it.

What's next for AIr - Your Personal Asthma Assistant

AI Chat Assistant: Get even more personalized guidance and tips based on your specific triggers and location, powered by OpenAI.
Wearables Data Tracking: Track your heart rate, body temperature, and exercise data to predict asthma-prone activities and help you avoid them.
On-Device Model Training: Enjoy enhanced privacy and security for your sensitive data, while receiving faster and more personalized insights.
Model Improvements: We'll continue to refine our model to include more potential triggers and allow for more detailed logging of your breathing troubles.