“Imagine a disaster where you need to find out where the survivors are,” said Willis. “You could have a fleet of little flying machines with the same kind of sensors that our animals have. Odour detectors. Vision detectors. CO2 detectors. Something that could fly into a building and locate a hot spot of carbon dioxide; that would be an indication of someone breathing.”

The military would love to have a robotic fly on the wall, he said. Why hasn’t this already happened? Some people have speculated that it has, hidden in classified military programmes. But scientists say we’re not even close to matching the skill and manoeuverability of a hawk moth, much less a little fruit fly.
For starters, batteries are heavy. It’s hard to find a source of energy that compares to insect muscle power. And the rules of aerodynamics say that small flying objects need to flap, and it’s challenging to make materials that have the flexibility and strength of an animal wing.

Finally, there’s the computation question: The robotic craft has to be able to perceive and react to its environment. It has to avoid flying into a wall. It may need to learn how to avoid being eaten. That is why scientists want to get inside the heads of insects. “The thing that everyone is trying to figure out is, what are the control algorithms that the brain is using to control this thing?” Willis said.

One possibility for the engineers: Find a way to piggyback on a flying creature. The military, Willis said, “would love it if they could robotically control a moth and have it fly into a window with a camera on its back.”
_Joel Achenbach, LATWP