WaterYouDoing

Do you ever stare at a slow growing plant and your just like "WHAT ARE YOU DOING?". Never again, as our all-in-one ecosystem ensures a suitable and fast-growing environment for all your plants.

Comment

Inspiration

We wanted to create an all-in-one plant ecosystem that records environmental data and tracks plant growth using computer vision. Our goal was to develop a system that could help plant owners monitor and optimize plant health with minimal manual intervention.

What it does

WaterYouDoing is a smart plant monitoring system that records:

  • Temperature
  • Moisture
  • Humidity
  • Plant growth tracking using computer vision markers

All collected data is stored in PostgreSQL and plotted on our frontend for easy visualization.

How we built it

We used an Arduino and the technologies listed in our orders:

  • Sensors: Humidity, soil moisture, temperature
  • Computer Vision: OpenCV's Open Library for contour mapping and object recognition.
  • Camera Feed: Connected via OpenCV to generate frames, mark pixels, and persist pixel height metadata to PostgreSQL.
  • Data Logging: Measurements were timestamped and written to PostgreSQL for plotting and linear regression to predict plant growth days.
  • Microcontroller Connection: PySerial streamed sensor readings to a lightweight Flask API, which wrote records to PostgreSQL. Backend and Storage: Flask + SQLAlchemy/psycopg with PostgreSQL for durable, indexed time-series telemetry.
  • Watering Mechanism: A servo motor was used for water pumping, with a backup wick system made of string in case the servo failed.

Challenges we ran into

  • Bluetooth module issues – It didn't connect properly to our devices.
  • Servo motor problems – Wires detached, making it unreliable.
  • Hardware transport – Moving from residence to the venue was challenging.
  • Backend-frontend integration – Learning how to connect everything in a short time.
  • Time constraints – Balancing hardware setup and software development was difficult.

Accomplishments that we're proud of

  • Successfully implementing a computer vision-based plant growth tracker.
  • Building a working prototype with multiple integrated sensors.
  • Overcoming hardware setbacks with creative solutions like the wick-based watering system.
  • Logging and predicting plant growth data using linear regression.

What we learned

  • Hardware debugging skills, especially with Bluetooth and servo motors.
  • OpenCV for plant tracking and contour mapping.
  • Frontend-backend integration and real-time data visualization.
  • Optimizing library usage for efficient processing.

What's next for WaterYouDoing

  • Scaling up with a peristaltic water pump for better water flow control.
  • Improving the housing materials for better durability and sensor accuracy.
  • Enhancing our computer vision model using Hugging Face to estimate plant depth automatically.
  • Optimizing the codebase for better efficiency and real-time performance.

We’re excited to continue improving WaterYouDoing and making plant monitoring smarter and more accessible!

Share this project:

Updates