Self-driving car

A simulator where you can interact with a self-driving car. The inspiration originates from the course Artificial Intelligence for Robotics - Programming a Robotic Car.

The car is using an algorithm called Hybrid A star to find the shortest path to the goal, which is the same algorithm used by the self-driving car Junior. You can read more about it here: Explaining the Hybrid A Star pathfinding algorithm for selfdriving cars.

The car is using a PID controller to follow that path and Reeds-Shepp Paths which Hybrid A* is using to increase the performance of the algorithm.

What are the lines meaning?

  • The gray and white lines are the shortest path if there were no obstacles. This path is the Reeds-Shepp path
  • The orange and blue lines are the Hybrid A star search tree (orange is if the car is driving forward and blue is reverse)
  • The red and light-blue lines are the final paths (one of them is smoothed)
  • The smaller white and black lines you can see along the final path are the waypoints the car is following. Black means that the car should move forward to reach the waypoint

Media

Instructions

Move camera:
WASD
Zoom in/out:
IO or Mouse wheel
Place the goal car:
Left Mouse
Rotate the goal car:
QE

Play it

Share it

twitter icon facebook icon

...and follow me on Twitter: @eriknordeus