Inspiration
What it does
How we built it
Challenges we ran into
Accomplishments that we're proud of
What we learned
Inspiration
Indoor spaces can often feel hot, uncomfortable, and polluted, yet most people have no simple way to understand their indoor air quality or know which plants would actually help. We wanted to turn the idea of using plants for healthier indoor spaces into something measurable and personalized.
This led to RootCool, an AI-powered solution that analyzes a room and recommends plants based on its size, conditions, and air-quality needs. Instead of simply asking users to “add more plants,” RootCool tries to answer the more useful question: which plants, how many, and where should they be placed?
What it does
RootCool allows users to upload photos of a room from four different corners. The system analyzes the images to understand the room and estimate its dimensions and volume.
It then generates a personalized RootCool Report, including:
- RootCool Score
- Outdoor AQI, temperature, humidity, and weather conditions
- Estimated Indoor AQI
- Estimated AQI after plant placement
- Number of plants needed
- Recommended plant species
- Suggested plant placement
- Estimated temperature reduction
- The most significant environmental issue detected
RootCool also includes a Plant Health Tracker, where users can upload weekly photos of their plants and use AI-based analysis to monitor their condition over time.
How we built it
RootCool was developed as an Android application using Kotlin and Jetpack Compose.
We integrated Google Gemini to analyze room images and generate structured environmental insights and plant recommendations. The application processes the AI response through dedicated parsers to convert the model's output into meaningful values and recommendations that can be displayed in the app.
The system combines computer vision-based room analysis, AI-generated recommendations, environmental data, and plant health monitoring into a single workflow.
Challenges we ran into
One of the biggest challenges was extracting reliable information about a room from only a few photographs. Lighting, camera angle, perspective, furniture, and incomplete visibility can all affect image-based estimation.
Another challenge was making AI-generated responses consistent enough to be used inside an application. Raw AI responses can vary in structure, so we had to design parsing and validation logic to transform the output into predictable data.
We also had to balance technical accuracy with a simple user experience. The goal was to make the technology work behind the scenes without making users understand how the underlying AI works.
Accomplishments that we're proud of
We are proud of turning a relatively simple idea, using plants to improve indoor environments, into an interactive AI-powered application.
RootCool does more than recommend plants. It creates a personalized environmental analysis, estimates the potential impact of plant placement, and provides a way to continue monitoring plant health after installation.
We are especially proud of combining room analysis, environmental data, AI recommendations, and plant tracking into one cohesive mobile experience.
What we learned
Building RootCool taught us that developing an AI-powered application is not just about integrating an AI model. The quality of the final product depends heavily on prompt design, structured outputs, validation, data handling, and user experience.
We also learned how challenging real-world image analysis can be, particularly when measurements need to be estimated from uncontrolled photographs.
Most importantly, we learned how to turn an open-ended environmental problem into a measurable, user-focused technology solution.
What's next for RootCool
Our next goal is to improve the accuracy of room dimension and plant-placement estimation using better computer vision and calibration techniques.
We also want to expand the plant database, improve personalized recommendations using factors such as room orientation and sunlight availability, and make plant-health monitoring more detailed.
In the future, RootCool could also integrate with IoT sensors to replace estimates with real-time indoor measurements and continuously adapt recommendations based on actual temperature, humidity, and air-quality conditions.
