Watermore | Devpost

Arduino Circuitry
Object detection and classification with confidence rating
Text sent from Twilio

Inspiration

This past year, I (Jemima) managed to kill three of my succulents. Arguably the hardest plants to kill. Keeping house plants has been proven to reduce stress, improve your mood, and performance. However, the busyness of daily life leads many of us to neglect our beloved plants and maybe even switch to plastic greenery :scream:. Without removing the responsibility of caring for your house plants WaterMore tells you when your plants need a drink!

What it does

This project is a detection system which utilizes sensors to determine when your plant needs to be watered! If you are within the view of the plant, it will simply yell to you reminding you to water them! If you are not in view of the plant, you will be sent a passive-aggressive SMS from your plant reminding you to water them ASAP

To do this it follows the following process:

Determines if the moisture content of the soil is below the desired threshold
If the plant needs to be watered: take a picture of the view of the plant
Process and Embed the image
Send image to Tensorflow model performing object detection
Objects in image are classified and confidence ratings are produced
If a person is identified in the image, the plant screams for a drink, otherwise you will receive an SMS reminder

How we built it

We used an Arduino Uno with a soil moisture sensor to detect the moisture content of your plant. The Arduino sends information to the pre-trained TensorFlow Machine Learning Model to detect and classify objects in the image. In the case where the plant needs water but no person is detected, the SMS is sent using the Twilio programmable messaging API.

Challenges we ran into

A main challenge we ran into was integrating both the individual hardware and software components together, as it was difficult to send information from the Arduino to the Python scripts we wanted to run. Also, there was a dependency issue relating to the Object Detection API from TensorFlow which made to visualize the object classifications.

Accomplishments that we're proud of

We are proud of successfully integrating both software and hardware components together to create a whole project. Additionally, it was all of our first times experimenting with new technology such as TensorFlow/Machine Learning, and working with an Arduino.