Randomly, one or more of my AI enemies’ movement will get stuck in this weird loop where it starts and stops repeatedly without actually moving anywhere.
From what I can tell, it appears to be constantly generating a partial path with the goal location just being where the agent is currently stood.
The behaviour tree and in-game AI debugging constantly swap between these two states:
Let’s take a few steps back and look at the logic path verbally:
If the Agent is NOT where it needs to be, it seeks a new location, moves there, and restarts the loop.
The question is where in the process is it reaching “I can’t do that” status, causing a failed moveTo.
It’s looking like the goal location is possibly too high to reach, causing a failed, retried moveTo without generating a new location? Do you want to check that out before we dive any deeper? Honestly the logic of the BT is sound.
Managed to get it sorted, the issue ended up being that the EQS was giving target locations that were jammed right into corners that the AI couldn’t quite get to. Thought the path exist filter would have covered that but it required a navigation projection instead.
Thanks you your input though, helped me start thinking and looking in the right places.
