Cropter

Inspiration

We were inspired by Blackberry Farms, a local farm in Cupertino that recently lost money due to pests at the very corner of their crops. Now, while this may have only been a small amount, for larger producers, pests are a very big threat.

What it does

The website allows the farmer to specify a single area that is their farmland on a global map, and then mark any no fly zones, which may have trees, buildings, etc. The backend then creates an automated flight plan to avoid these no fly zones, but also map the entire field, marking any pests, or other problems; later this flight is analyzed and quantified for the farmer. Should the farmer want to just go around and monitor the crops, he may use WASD and the keyboard to move the drone around and monitor.

How we built it

We built it by breaking it down into 4 phases; the initial setup which gets the field's area and diagram, and where to fly, this we did through Leaflet.js. Then, in the 2nd phase, we created an automated flight plan for the drone to follow to examine the entire field. In the 3rd phase, the 3 models analyze the footage, and replicate the video, except with markers for pests diseases, damage, etc. Finally, on the farmer's dashboard, the stats are shown for the analysis, like the area of good crops, bad crops, the estimated yield based on the crops that will be fine for selling, etc, this is done through another model that finds patterns in the flights and quantifies them.

Challenges we ran into

A challenge we ran into was the Wi-Fi as it slowed down the download of some very large datasets and packages. Additionally, finding a viable dataset was hard as there weren't many that were the size we needed and had the images we needed.

Accomplishments that we're proud of

We're proud of the models we trained and their accuracy, as well as the actual flight automation and how it creates a flight path.

What we learned

We learned how to improvise, and how to work through challenges because, originally, we planned to do too much and it was simply too hard to implement so much. Thus, we chose the most important aspects and focused on improving those.

What's next for Cropter

We'd like to try to use our website with a drone that has a bigger battery life and a farther range to expand. Also, if possible, we'd like to add attachments that work alongside the drone to deliver resources like water, fertilizer, etc.

Built With

• leaflet-js
• python
• react
• tailwind
• typescript
• yolov8