Re: ants again

From: Anders Sandberg (
Date: Tue Oct 23 2001 - 04:37:41 MDT

ObExtro: ants are a great example of spontaneous orders and multiagent
systems. If we can learn how they do what they do, we can build
efficient microrobot and nanodevice swarms for many purposes. Besides,
we are likely to have ants around for a long time in our ecosystems, and
they might be a cheap way of doing micromanagement.

On Mon, Oct 22, 2001 at 09:18:41PM -0700, Spike Jones wrote:
> My notion is that the ant migration I witnessed yesterday was
> started by an ant that found a suitable place for moving the
> colony, so she walked back to the colony trailing a chemical which
> means in ant, "everyone pick up a larva and walk this way."
> So here's the idea. If I can find another ant migration in progress,
> I attempt to recover and produce whatever chemical means
> "grab an egg and come". Then whenever a colony of ants is
> discovered in the house, one simply brushes a strip of the chemical
> from there, leading to the outdoors. Then the colony picks up
> and walks right outta your house, carrying the young with them.

Sounds good. I have always wanted an ant regulation kit.

I'm not sure how they detect the pheromones, but there are several
research groups studying ant pheromones that might have good ideas or
techniques. C.f.

Here is one abstract I found:

Putative Alarm Pheromones of the Ant Species Formica obscuripes
(Hymenoptera, Formicidae)
Warren J. Wood

Alarm pheromones of the ant species Formica obscuripes were
investigated. Volatile compounds in the headspace above aggravated
worker ants were collected by solid-phase microextraction (SPME) and
analyzed using gas chromatograph-mass spectrometry (GC-MS). Analysis
revealed the presence of the suspected alarm pheromones decane,
undecane, tridecane, 4-tridecene, pentadecane, and heptadecane, as well
as several unidentified components. The identities of the
straight-chain hydrocarbons were confirmed by comparison with mass
spectra of authentic samples. The location of the double bond in
4-tridecene was determined by a standard methylthiolation derivatization

Seems to be simple chromatography.

> So I need to create an instrument which would recognize the
> difference between a zero-flux ant-line and a true one way
> migration like the one I saw yesterday. Wouldnt that work?

Sounds like a somewhat tricky problem - not that I think computer vision
isn't up to it, but it would need training data.
> I even had an inspiration on how to recover some of that
> chemical. When I watched the migration I noticed that the
> forward antflux was about 10/second, and the reverse
> had about 1/second or perhaps 0.5/second. So what were
> the backwards goers doing? I figure they must be "repeaters",
> gals going back along the line replenishing the scent trail which
> would otherwise evaporate. Does that sound right? They
> would walk backwards along the trail, excreting a new line
> of come-this-way-juice. What else would they be doing?

The documentary I watched demonstrated the same thing among army ants,
where the repeaters did rush back to show others the way.

> So when I find a migration, I need to capture the backwards
> walkers and find out what they are excreting. Right? spike

Suck air above them, gather into a chromatograph sample and analyse. But
don't disturb them, or you get the alarm pheromone instead.

Anders Sandberg                                      Towards Ascension!                  
GCS/M/S/O d++ -p+ c++++ !l u+ e++ m++ s+/+ n--- h+/* f+ g+ w++ t+ r+ !y

This archive was generated by hypermail 2b30 : Sat May 11 2002 - 17:44:15 MDT