Many research traps are permanent, but not very obvious.
Operating a trap in the same location over many years has some strong advantages in providing good statistical data. There is a network of suction traps across the UK co-ordinated by Rothampstead which serve to monitor aphid populations (and incidentally other aerial insects). Many sites operate permanent light traps for monitoring moth populations (e.g., museums, research stations etc.). There are also likely to be permanent Malaise traps. The common factor is that these are not necessarily very robust & therefore are likely to be sited in areas less accessible to the public.
Note that orchards and other crop sites may operate permanent pheromone lures/traps which look like little bits of cardboard with a sticky surface.
Personally I don’t see why animal_trap cannot be used, with appropriate sub-tagging (there are multiple types of bug, mammal & bird traps anyway).