This reasearch is spot on to what I've been saying about traditional AI for a couple decades now. Insects fly thru a 3-D environment avoiding static obstacles and moving threats with vision and cognition systems simple by our current Megapixel and Intel Inside capabilities. These researchers are approacing the problem utilizing concepts stolen from evolution and applying a simpler, different approach than the university lab groupthink. Good for them. A synergistic approach combining std computing with simple ideas from simple life could result in robust real world capable systems.