"Smellicopter" uses amputated moth antenna to navigate towards smells

Originally published at: https://boingboing.net/2020/12/10/smellicopter-uses-amputated-moth-antenna-to-navigate-towards-smells.html


Your biological and technological distinctiveness will be added to our own. Resistance is futile.


Did they give the moth cybernetic replacement?


This is the culmination of several labs research. A Japanese group first used a whole moth on a wheeled robot a decade ago to feed the antenogram into a µC for automated odor tracking. Since there have been proof of concept studies showing that a robot can learn the signature of explosive scents to train it to track undetonated mines.

The wire going into the antenna is a recording electrode; it measures the currents generated by the antennal nerve. They aren’t sending any current into the antenna.

On a technical note, if they switched to a suction electrode (which would prevent the nerve from drying out) they could keep the antenna alive and functioning considerably longer than 4 hours.


imagine the results if they put a dog’s nose on it instead.


I assume your joking, but I’ll respond anyway. Pardon the horror of animal research

Mammals and other verts have closed circulatory systems that pressurize the blood to get oxygen throughout the body. When you start cutting off body parts, then blood loss makes death of the animal a big risk and the amutated part dies because it no longer gets oxygen or energy. In insects, there is an open, unpressurized circulatory system and passive gas exchange. The animal doesn’t bleed out and the amputated part keeps on trucking pretty well as long as it doesn’t dry out.

Also, foraging insects are master smellers. There’s no reason to assume a dog’s nose would be more sensitive

1 Like

“To get a signal out of it they just run an electrical current though it.”



But the moth antennae wouldn’t sniff out your lost frisbee

I remember seeing this at a UW science fair a few years ago, it was incredibly cool and the grad student researcher who was presenting it was very understated. The main theme of the event was rocketry and many other cool projects were on display, but the the zombie moth chemosensor really stuck with me as it was so totally outside of my expectations.

1 Like

Do you know if the antenna structure at a multicellular scale is critical to the arrangement working?

I’d think that, for pure convenience, the ideal would be one of the cell types that dehydrates and rehydrates readily so that you could store it in inactivated form and then mix and apply when needed(since there wouldn’t be a proper nerve to jack into presumably you’d use a very finely spaced contact array instead of a couple of large electrodes); but if odor sensitivity is a multicellular business, not something a cell can do on its own, that would obviously make any more or less homogeneous EZ-mix cell slurry less useful, even if you had a candidate that rehydrates well and has the necessary characteristics to be a nerve cell.

I wonder if the nerve-stump of the moth is sensitive enough to low voltages that you could swap in a literal antenna without any active amplification or similar tricks: just take a conductor of the right length and you will get some electrical activity out of it when exposed to RF; so given a sufficiently sensitive nerve or a sufficiently punchy radio you could presumably give the moth creepy radio wave sensing powers with just some really thin wire.

It’s the one endlessly circling the wifi access point rather than the lightbulb, that’s how you can tell.


When a neuron dies, it can’t be revived. When I say the nerve needs to be kept from drying out, I’m over simplifying some. A fairly precise balance of salts within a narrow concentration range needs to be maintained. Once their gone, thats it. So no way to do an instant mix or even careful cell culture to reproduce it.

The moth antenna has ~300,000 olfactory neurons with a very complex physical structure to control airflow over the receptors:

The antennal nerve has thousands of small (< 1 µm each) axons projecting to the antennal lobe. It only takes mV,mA signals to stimulate action potentials in them that would then go to the brain. Sending a single signal to the whole nerve, though, would generate only the crudest of percepts. Maybe like an optometrist putting a bright light a centimeter from your eye. If you somehow could actually attach an external antenna that activated enough of the axons independently based on the signal, the moth could absolutely learn to discriminate whatever the sensors were detecting

1 Like

No one but me is curious as to WHAT the drone smells? Something interesting to a moth, I’m guessing.

The poor moth, flying around with no “nose”. How does he smell?

What kind of odor were they testing for? Female moth pheromones don’t have much lifespan or other use (except for the obvious). Can they reprogram the antenna sensors to find explosives or maybe an individual?

So how long till we have factory growing moths to amputate, for our insatiable demand for moth parts and smell copters, and if it only last 4 hours that a lot of moths, and do we have to learn how to butcher our own moths?

OK, I’ll bite:


Ba Dum Tish!

This topic was automatically closed after 5 days. New replies are no longer allowed.