I had an awesome time contributing to TerraMoist, a smart soil moisture monitoring system designed to help farmers optimize irrigation. Here’s a breakdown of my work: What I Worked On Hardware Hacking & Sensor Integration
Prototyped the initial sensor array using capacitive soil moisture sensors (no corrosion!) and ESP32 microcontrollers.
Fought through ADC noise issues (those analog signals are finnicky) and implemented software smoothing.
Helped design the low-power sleep mode to extend battery life in remote deployments.
Backend & API Shenanigans
Built the Flask backend to ingest sensor data and store it in a time-series database (InfluxDB).
Integrated Twilio for SMS alerts when soil moisture dropped below thresholds. Farmers loved this!
Wrestled with MQTT for real-time updates (retained messages FTW).
Frontend Dashboard (React + Mapbox)
Developed a live-updating dashboard showing moisture levels across different fields.
Added geofencing to visualize dry zones super useful for large farms.
What I’m Most Proud Of Making It Actually Useful – Seeing farmers rely on our system to prevent over/under-watering was incredibly rewarding.
Battery Life Wins – Got our field units running for 6+ months on a single charge by optimizing deep sleep cycles.
The Alert System – A farmer once told me, "This text message just saved my crop."
What I Learned Hardware-Software Glue is Hard – Sensor calibration, power management, and wireless reliability are brutal but crucial.
Farmers Care About UX Too – If it’s not simple, they won’t use it. We iterated a lot on the alerting logic.
MQTT > REST for IoT – Way lighter for frequent small payloads.
Log in or sign up for Devpost to join the conversation.