I have a game idea where I need to know beforehand how actors would navigate a given environment. Is there a way to invoke the navigation/pathfinding system without doing it in real time? In the end I want a ’trail’ of points that the system would use to get from A to B without this being solved in real time.
Hey @x_jdj_x, welcome to the forums!
So we don’t have any OFFICIAL documentation on this, this is a really good video of how to go about what you’re proposing. If you want to copy what he’s doing exactly, you’ll need to start the series from vid 1, but it’s great for doing exactly what you are wanting, I think. 
Disclaimer: This link is unaffiliated with Epic Games or its partners. Please click at your own risk.
1 Like
Thank you for your reply, however it was not really what I had in mind; the video explains how to navigate using a sequence of waypoints.
What I had in mind was to create a point trail (more or less on a frame by frame resolution) before the actual play start and then during the playthrough, use this sequence as the positions during actual gameplay.
So my thinking is in line with doing a compute loop using the navmesh to solve actor movement before the actual gameplay and then replay it.
1 Like
Is this what you need?
It will output the navmesh calculations as a vector array that you can save and then reuse.
It’s between two actors but if you have the points as direct vectors then just plug them in directly.
The timer function is just to update it while dragging in the viewport to see it in action (only the function is needed)
4 Likes
This indeed looks like what I am looking for! I will investigate it later, but I will mark it as the solution as I’m quite convinced it is!
Thank you!