My public personal notebook – Not meant as a guide
Brushed DC motors are cheap and robust, but inherently hard to control accurately compared to modern BLDCs, especially at low speeds. This is fine for applications like household appliances or electric power tools, where positioning and precise speed control aren’t needed. For a skid steer robot, however, it is a prerequisite even for just driving…
Those little Arduino robot-car kits are neat and all, but sometimes something that provides a bit more space for electronics, something that has a little more OMPH just sounds like much more fun! So let’s build a box! With wheels and motors and all, obviously. Let’s call it a rover. Here is the wishlist for…
To train a driving model, we need driving data. Let’s break down what exactly we’re interested in. A fundamental requirement: We don’t want to do any manual labeling. We will need at least a couple hours of driving, which means 100.000s of frames recorded. We can’t possibly label those by hand within the scope of…
We want our camera-based system to run on an nVidia Jetson Nano, display a GUI on a monitor and talk to the car via a CAN-USB-interface. It should know its GPS position and be able to turn the steering wheel using a stepper motor. Guess what my car is missing? All of that. Let’s add…
Yeah, yeah, I know. Who needs another article about training some DNN on a bite-sized Kaggle dataset, and then never do anything with it, right?
That’s not what this is about. We’ll put this is a real car.
Here are the steps.
