Local avoidance precision

General
1

Hi, I’m including the local avoidance system into my code, but there are things I would like to ask regarding the results I’ve had so far.

I’ve created a test situation where one cylinder is trying to reach a destination that stays directly behind another cylinder that is close to a wall (Screen 1, lower part of the scene; I’m using NavMesh in it). The cylinder starts moving towards the wall, since it’s the shortest path to the destination, but gets stuck between the other cylinder and the wall (Screen 2). Then, five seconds later, it finally moves around the other side (Screen 3) and move towards the destination, but once it gets there, it keeps walking around the destination point and doesn’t stop moving (Screens 4 and 5).

So, here are my questions:

  1. Is there a way of making the agent finding the alternative route faster? Inside a combat scene, five seconds could be an eternity.

  2. Is there a way of making it stop at the destination point without increasing the End Reached Distance parameter? The system I’m creating demands that the destination point should be as precise as possible.

Here are the screenshots:

Screen 1:

Screen 2:

Screen 3:

Screen 4:

Screen 5:

Despite that, the overall system is great (much better than Unity’s), so congratulations on the great work! :slight_smile:

Thanks in advance and best regards,
Edgard

2

For some reason the screenshots appear as broken, but if you right click on it and tell them to open into another tab, they work fine… :frowning:

3

As a follow-up, I was able to solve the problem partially by switching the RichAI class by the AIPath class – now the actor is able to reach the destination point properly, but the AIPath class doesn’t have the wall force parameter, and I already caught one situation where AIPath doesn’t respect the navmesh walls.

All in all, looks like it’s something related to the RichAI class, so any ideas in this matter?