With a decent (perhaps high speed?) camera and a trained computer vision model, couldn't a contactless study be performed of flying insects? It should be possible to count the common species in an area with reasonable accuracy and flag photographs of unknown species for human review. Definitely not as cheap as a trap, but perhaps with solar power could be just as easy to place. Unless someone with experience with these studies can make the case that the number of insects killed by these traps is negligible and all this effort isn't worth it of course.