SuperSeed

Inspiration

As society has shifted further away from agrarian lifestyles, people have become decreasingly familiar with the work that goes into growing plants. While everyone may recognize that soil is important for agriculture, people may not be aware that inspecting the quality and conditions of the soil is highly important for growing crops and other types of plants. Typically, this is conducted by taking samples of the soil and sending it to a lab, but what if there was an easier way...

What it does

Our SuperSeed will contain several sensors for moisture, temperature, conductivity, organic matter, and soil compaction. Using these inputs, our SuperSeed will be able to predict pH, nitrogen, potassium, and bulk density properties for the soil. The SuperSeed will be placed in the soil and will be able to relay these predictions to an app.

How we built it

Due to our minimal experience with machine learning, we used AI to give us the skeleton for our code, then we filled in the specific parts for this project while building upon the foundation. Furthermore, we theorized a potential design for the physical project which incorporates the use of sensors and a radio to collect the data and return it to a server.

Challenges we ran into

As this was our first time experimenting with machine learning, the most challenging part of this project was understanding how to implement the machine learning using environments and dataset aspects of the code to our IDE. Additionally, we wanted to make a realistic model for what the product could look like, and due to its size, we ran into several limitation when trying to locate the correct parts for the device.

Accomplishments that we're proud of

Using the starter code we acquired, we were successfully able to write code that could theoretically take data from a sensor and predict the conditions of the soil based on the observed data. With this being our first time using machine learning, we were very pleased that we were able to create a functioning code.

What we learned

We learned how to effective implement machine learning environments and datasets into our code. We learned how to train the machine using given data and then predict outcomes from it.

Built With

• anaconda
• pandas
• python
• scikit-learn