Basic logic for this is quite simple. Transform the world 2D location of the actors you wanna show in the radar, to the radar.
The radar would be in UMG, and it’s 0.0 is on the top left of the screen.
Say an actor is at “X = -100, Y = 500”, just divide that, say by 100, and now they have a -1, 5 coordinates. which you can use to place a small image on its relative coordinates. You would just have to offset the scaled coordinates by the location of the radar on the screen.
Now, bear in mind that of all things, a radar is actually pretty expensive on the CPU and you should most definitely do the computing parts on C++ on the long rung, unless you start with a “lazy” radar that doesn’t do everything on a single frame, but you could still get a noticeable frame drop for that. You still do the visual parts on UMG.