I have a 3d environment with a flat plane where the user can freely move obstacles. The AI should still be able to find the shortest path after the environment has been changed by the user.
Any advice on how to approach this?
Thanks 
You can try to split the field into cells, and find the way with the A* algorithm. But you will have to update the cells often
Here is a good example
I think you can use our navigation mesh which has support for adding/removing obstacles. @Cedric can correct me when he is back.
2 Likes
Yes did a test, and the webworker/create navmesh is such a great feature, think this will do the trick. Thanks a bunch
1 Like
I experimented with Threejs & Yuka before i got here. Yuka has many great features but there is no built in navmesh generation as far as I know.