Introduction

In mid-2021, as I improved my software engineering skills I wanted to take on a new challenge. For most of my life I have had an interest in remote vehicles and software/hardware integration. I finally decided to commit and purchased an Nvidia Jetson Nano 4GB and remote-control car from Amazon.

I found an online community that has been working on self-driving RC cars for a while and I used their DonkeyCar software to get started.

Fitting The Pieces Together

I had the Jetson Nano, RC car, and a number of recommended parts from the DonkeyCar website. I wanted to be able to use the Jetson in the 4GB - 10 Watt configuration which required at least a 5V 2A power supply. I had a 2-cell LiPo battery lying around from an RC glider and I bought a DC-DC buck converter to step down the voltage from the battery.

I picked up another small 5V NiMH pack to power the PCA9685 PWM servo driver used for the steering servo. This helped to balance the power draw and protect the PCI bus on the Jetson. I also picked up a bigger LiPo and the IMX219 camera and wired everything up.

In order to fit all of the components on the RC car, I used some scrap acrylic and drilled holes to fit it on the chassis body posts. After some gorilla glue, zip ties, and velcro, everything came together into a somewhat ‘neat’ package.

Final Product

Many hours of testing and debugging later, I was able to record, train, and deploy a model to the Jetson. The debugging was the most time intensive part of the process. Because of the Jetson Nano’s small footprint in computation and power draw, there are some features within the Nvidia JetPack SDK that prevent flexability among certain Python libraries. Thankfully, the DonkeyCar Discord server was responsive to bugs that would pop up, both for myself and many others using the framework.

I went down to Circuit Launch in Oakland for a racing event a few weeks later and had a chance to network and compare builds with some of the community.

This was an incredibly fun project to work on and I will definitely be pursuing more drone projects in the future. Navigating 3D space is how we traverse and communicate with the world, and I’m excited to dive further into this problem!

I’ve learned a tremendus amount from this project and happy to share more with whoever is interested! Thanks for reading!