what about using vision / collision cones to adjust the torque applied? and then use way-points with a speed variable that each car adjusts based on it’s skill and handling etc,
so “Car” cone left hits waypoint, adjusts torque or steering to line up with new way point(or close enough) and then when it drives through, the next point loads…
this way each waypoint can change etc based on changes in the map or if someone is on a waypoint,
another issue would be to have lots of “safe” zones and then use A* to decide the path…
or a reg nav mesh, or another option is to use a curve as a “target” that the car attempts to run over, and the closer it gets, the more the car brakes? so the “rabbit” for each car is a decent run that it bases it’s motions on?
these cones could also engage breaking if there is a car not moving away fast enough in the collision path…
I do plan on creating an advanced racing AI tutorial, though I may do an Intermediate as well.
The Advanced tutorial would use loads of python and probably a waypoint/ray/navmesh hybrid style AI. Whereas the intermediate would be easier to follow and use less python, it would also be less “Smart”.
Thanks for the feedback, Its what helps to improve tutorials!
Thx for the tut
Learned a lot…u are going a bit fast but after watching it twice i got it (or maybe i am very slow at learning LOL)
Please do make some more tut’s about AI