OptiGrow

Ensuring the optimal growth of your plants even in a drought. This product detects the plant species and the moisture level of your soil and waters it until the necessary moisture amount is met.

Comment

Inspiration

With the frequency of climate disasters occurring as a result of global warming, droughts are becoming more common across the country. One of our members, as a California native, has brought up their desire to keep plants in their home while not wasting the precious resource of water in the process. Realizing this may be a desire among many people, we decided to develop a way to efficiently water plants based on the species and the average moisture levels needed for them to survive.

What it does

Using AI, the product recognizes the flower/plant species and calculates the optimal soil moisture necessary for it to survive. By using this as a threshold value, the moisture sensor tracks the moisture level every hour, and waters the plant if below that threshold.

How we built it

We utilized Arduino and a water pump for the hardware aspect to control the watering, as well as a SeeSaw moisture sensor for moisture detection. In terms of software, we used TensorFlow (running on Google Collab) to upload pictures of the plant and get image recognition, and then python/Arduino software to use the plant type to calculate the threshold for moisture and set the loop for watering it based on that threshold.

Challenges we ran into

  • Image formatting: It was difficult to take an image of the plant we have and upload it to the database in a format that is automated.

Accomplishments that we're proud of

  • Trained an AI model to recognize flowers and various other plant types.
  • Python to Arduino communication for watering and moisture sensing system.

What we learned

  • How to communicate with Arduino using Python through the serial port
  • How to create an image classification AI through TensorFlow using Google Collab

What's next for OptiGrow

  • Streamlining image capture and automatic input for plant type
  • Implementing app/IoT to allow users to manually water and monitor plant health (possibly using Blynk IoT or Adafruit software)
  • Improved material, such as acrylic or injection molded parts to provide greater product stability

Built With

Share this project:

Updates