It is said that in a few years the automotive industry will produce cars with more software than hardware. In the Autonomous driving sector it’s easy to see that this is already a reality.

We will see an infrastructure of different libraries and technologies working together for solving the task of making a car achieve autonomous driving. From the Vision sector with the task of perceiving the obstacles, the trajectory path planning that calculate the path using different logics, requirements and objectives. Finally converting these scientific calculations into practical controls for the car and then integrating them in the car system.

We’ll see in detail how the use of libraries like Numpy and Matplotlib are common in this context, and how other scientific libraries like scikit-learn make things easier and faster: critical when you have to work with expensive Formula racing cars.

These tools have their limits, and it’s interesting to test them for having a deeper understanding of every technology put into the context of autonomous driving; overcoming these obstacles is crucial for optimizing safety and performance.