Smart Tractor

Still shot of the robot
Mission Control
GPS Map
Object Detection of Person and Dog
Rover Working

Inspiration

We really just wanted to build a mini monster truck and tried to figure out a way to spin it so it makes sense for a hackathon project.

What it does

It collects a bunch of different kinds of data and makes use of a smart camera that leverages different data sets for image detection. The robot can drive around to designated gps coordinates on our website. When it arrives at the target location it will sample data using its various sensors like soil moisture, temperature, air quality, and humidity. While it is driving it can be configured to detect different animals using datasets like birds which can then be configured to drive towards them to scare them away.

How we built it

Basically the robot will receive GPS coordinates in JSON format from a web server built using Express.js. The GPS coordinates are set by the web interface built using Google Maps and React. The robot will then adjust its position to drive towards the new location. When it reaches the destination the rover will sample its environment and uploads its findings to AWS using our web server. The web server will save those findings to a MySQL database.

Challenges we ran into

Getting our hardware to work. We couldn't get our original camera working so we had to switch our focus around to get something that worked.

Accomplishments that we're proud of

We're pretty proud of getting the robot actually driving. We pretty much had no idea how this robot operated before and it took some time just getting the wheel to respond as our driver was not working.

What we learned

We learned that building a complete robot from scratch is probably not feasible in a 24 hours time frame.

What's next for Smart Tractor

We plan on improving the basic functionality for the robot such that it can run and drive effectively, right now it's very barebones just to make the hackathon deadline.