SOILution

Inspiration

Farmers across the Prairies are losing crops and income because of droughts and unpredictable rainfall. Many rely on fixed irrigation schedules that waste water and fail to meet each crop’s real needs. We wanted to build something that gives farmers control during uncertainty, a system that reads the soil, learns from it, and helps them act before damage happens. SOILution was born from the idea that data can make every drop of water count.

What it does

SOILution monitors soil moisture in real time through ESP32 circuit boards placed across different zones of a farm. It keeps soil moisture within healthy thresholds set for each crop type and automatically activates sprinklers when levels fall too low. The system also evaluates soil health by tracking how the soil responds after watering, giving the farmer a confidence score that reflects the risk of spoilage. All of this is displayed on a clean web dashboard that shows live data, trends, and water savings.

How we built it

We used ESP32 boards with soil moisture sensors to collect real-time readings and send them to a Supabase cloud database via Wi-Fi. The dashboard was built using JavaScript and Vite, allowing farmers to log in, view their farm zones, and visualize moisture trends using interactive graphs. The backend stores crop-specific thresholds, analyzes patterns, and triggers irrigation automation when needed.

Challenges we ran into

We faced issues with sensor noise, data stability, and synchronization between the hardware and the Supabase database. Calibrating thresholds for different soil types and crops was tricky, especially when there are not that many easily accessible resources available about the soil moisture thresholds for each type of crop.

Accomplishments that we're proud of

We’re proud that SOILution brings automation, soil health analysis, and water optimization together in one system. Our model doesn’t just detect dryness, it learns how soil behaves and adapts irrigation intelligently. Seeing the dashboard show real-time moisture levels and automatic responses for the first time was a huge moment for us.

What we learned

We learnt how to connect many hardware to the same database and learnt to create modular database schemas. We also learnt crop behaviour and how to address each soil condition to bring it back to its optimal conditions. We also realized how powerful IoT and data analytics can be when used to solve real-world environmental problems. Most importantly, we learned how to design for impact not just function.

What's next for SOILution

We plan to improve the predictive side of the system by training models that anticipate drought patterns and recommend proactive watering schedules. We also want to integrate weather forecasts and satellite data to make our insights even more accurate. In the future, we aim expand by partnering with Prairie farmers and track real-world water savings across multiple crop types.