MicroCART History

MicroCART (Microproccessor controlled aerial robotics team) has been around since 1998. In the past, the MicroCART worked on developing their own large-scale drones from the ground up. Several of the previous prototypes can be seen in the Coover 3050 lab. For the last two years at least, one of the primary goals for MicroCART has been to create, support and update the MP4 lab for CPRE 488 which has to do with crazyflies.

2022 - 2023 MicroCART Team

Our task this year was to continue development for the CPRE 488 MP4 lab. We wanted to take the work that the previous years team did, and to iterate on it so that students would have less headaches while completing the lab. The largest changes we made to the lab were completely restructuring the lab document to feature both instructions and appendixes rather than too much explanation all at once. The other large change was to the GUI, we added a "Update Params from JSON file" button that uploaded parameters to the drone all at once, which saved a lot from random crashes the backend would experience due to sending parameters too quickly. We also made it so the frontend connects to the backend automatically. Another large change implemented by our team was recording a few explanation videos for MP4 part 1 and part 2 which would be intended to be watched by 488 students that can guide them through how to use the software for both parts of the lab and some helpful tips and tricks on how to make progress.

Additionally, another key task for our team was to develop our own quadcopter. Developing quadcopters is a common DIY project, but the goal for our project was to make a platform capable of being a research test bed. In order to do this, the goal was to make a raspberry pi 2w be the brains of the quadcopter, and distribute power using brushless motors and an ESC. The software for our drone will be a somewhat modified version of the crazyflie firmware that interfaces over I2C with our ESC and sensors, and over wifi with the crazyflie CRTP protocol.

An effort that was mostly kept on the backburner this year was developing a quadcopter model simulink and matlab model. This model is able to use either step input blocks, or gamepad input in order to control the setpoint for the drone. A PID and LQR version exist. Most of the hard part is all from some example code Jones gave us at the beginning of the year. We wanted to take this model and try getting it to the point that you could basically complete MP4 part 1 in the model, but never got to the point that we could test if this would work.

The final component of our project was to develop an impressive demo to showcase to incoming students during scholars day.

If anyone needs to contact us for anything please let us know we are all super passionate about this project and want to see it succeed!

If you can't get in touch with anyone bug Grant or Tyler and they will get you what you need.