Trying (and so far failing) to reproduce the perception system for pedestrians at a crosswalk as demo'd in part 4 of last year's academy sessions.
One possible reason is that I am totally unclear about the correct approach to scaling the perception cylinder.
My environment units are centimeters and my entities are being scaled up x100.
So, should the cylinder be scaled to the source entity (this is what I'd expect intuitively, which scaling happening automatically), or scaled up 100 to match character size in the scene? I've tested both approaches but am still confused as I've seen Golaem display the perception volume both correctly and incorrectly scaled.
Also, it it necessary to freeze transforms on perception objects?
So far I've not been able to get the perception system to "see" any of the external entities I have set up in my scene, even though everything seems to be wired correctly.