SmartPlantAlive?

• 

  Poster of our idea

• 

Inspiration

We (AIoT Fusion) have tried growing plants at home before, but it failed after a couple of days. Making planting accessible to all is our goal. We aim to win with a sustainable project.

What it does

SmartFarmAlive is an AI-enabled IOT solution with unique features:

• Automated water sprinkle enabled when water sensor detect low level of water
• Light that will adjust according to the light sensor level
• Letting owner know plant survival using AI to predict if the plant is going to survive
• Using image recognition to determine leaf's colour

How we built it

Using Arduino to construct the hardware required to collect data such as plant imagery, moisture level and light level. We used the imagery collected to determine the colour of plant's leave using Azure Computer Vision. Based on the light level, moisture level, colour of plant, number of days and height of plant, we trained an ANN (Artificial Neural Network) to train a predictive model to predict how likely our plant will survive. Another feature that we have is a water sprinkle. When the moisture level is low, our DC motor will run and water the plant. Main languages utilised is C++ and Python.

Challenges we ran into

As this is the first time we are dealing with Arduino and AI, we had many challenges to get the motor and light running. AI was confusing to us as we did not know what model to use and how much data we need to train.

Accomplishments that we're proud of

We had a running prototype for both hardware and software.

What we learned

We work better in a team. We learned how to do Arduino which was outside of our course scope.

What's next for SmartPlantAlive?

Making it more seamless as our current prototype is not seamless.

Built With

• arduino
• azure
• c++
• python