Ants: a parallel

Ants like humans operate in complex ways, both in a collective and individual basis. Much research on ant behaviour has been carried out which shows parallels of ant colonies and various aspects of human behaviour, at a low level of cognition ants can be related to modern society. "Ant castes are exceptionally well suited for optimisation studies in socio-biology. ... individual worker ants are full organisms with ordinary, whole patterns of social behaviour, yet they are also clearly specialised for particular well-defined tasks...”[1]. They will overcome complex problems of locating and location by using pheromones and geometric recognition to recognise their position . "An ant colony is a complex system of individuals inter-acting with each other and their environment. Even though colonies have at least one queen, and they appear to act co-operatively and purposefully, there is no leader. Aggregate colony behaviour emerges from chemical cues, contact between individuals and environmental pressures. " [2].

Ants natural drive to wander and build indeterminate tunnels that create intrinsically interesting structures was explored in Yukinori Yanagi's 1994 'Asia Pacific Ant Farm' [3]. Yukinori constructed rectangular boxes joined by perspex tubing and filled them with coloured sands in the motif of various flags. Ants were then released into these where they dug tunnels through the flags which was used as an "allegory" to the visual element. The laboring ants provides a metaphor also to the modern way of life in a nation such as Australia. People have a direction as to a task in a seemingly chaotic manner which is a part of improving the whole (for right or wrong). In these fast passed processes little attention is paid to other beings except for a quick handshake or information exchange. With 'Eight-Bit Ant Farm' [4], Remo Campopiano 2002, ants were used in the installation as a chaotic object that was processed into data with lights and ping-pong balls affected by the ants movement. Data from the people moving around the room was used with the ants movements combing the two into a new environment data set . Remo and Jonathan Schull were using ants as, what they feel, "semi-autonomous" "seemingly random" material which the installation interpreted the chaos in a visual manner. Although ants seemingly act with no free will they are each individuals that take their own initiative in determining what to do and are able to survive without the colony system.

Audio Ants

Sound of the physical movement of ants and their visual presence has been widely used in inter-media works such as film [5], video installation [6] and sculptural installation [7]. In most common assurances a similar motif is adhered where the output of the ants physical presence is maximal. In the opening scene to David Lynch's "Blue Velvet" [5] a man has an aneurism where the camera then slowly zooms in on the ground revealing a growing visual and audio presence of insects crawling. This builds in intensity to the point of monstrous. Richard Lerman's video [6] of a swarm of desert ants crawling across microphones at high gain reversed the micro/macro existence of the insects. A more subtle implementation was inserted to Steve Peter's installation "Here-ings: a sonic geohistory" [7]. A number (12) of speakers were suspended in a space where environmental sounds of Mountainair, New Mexico were broadcast to recreate an environment which people could move freely about.
Implementation of augmenting ant colonies as a live social habitat has been explored extensively in scientific studies for creating solutions to emergent phenomena [8], human cognition of navigation [9],distributed systems [10], social networks [11] and reactive systems [12] to name a few. This plethora of human and/or machine paradigm's creates interesting avenues for extrapolation. Although ants have proven to be deaf they produce a communication sound called "stridulation" [13]. This is a near field sonic communication which is involves the air particle oscillation around the sound source. This means this is achieved by a number of cuticular edges rubbed against a scrapper [14]. The basic frequency for this is around 700Hz [13]. An excerpt from Flavio Roces and Jurgen Tautz paper "Ants are deaf" [15] demonstrates the method of sound generation as mono-pole, i.e. a pulsating sphere. The equation for mono-pole sound generation:
k * å < 1 (k=2?/?, å = radius of sound source, ? = wavelength).
Taking for the ant gaster (from where the sound is being radiated) a radius of 1-mm, and being 300-mm at 1 kHz, the conditions for a mono-pole are fulfilled. The near-field particle displacement amplitude drops from a zeroth-order sound source with increasing distance following
1/r2 (r = distance to the centre of the sphere);
from a first-order sound source with 1/r 3 , etc. The near- field particle displacement amplitude (d) for a zeroth-order sound source can be calculated according to
d = å2 * ? d / r2
where ? d = displacement amplitude of the surface of the sphere.

Following the Little Guy.

Object tracking systems are used as one of the fundamentals in computer-vision systems when deployed for real-world intergration. A number of methods can be employed that give stable results, colour difference and movement detection are suited for such a task just as they are for the humans. Keeping track of multiple agents in space/time has proven difficult when each has no defining marker, or when two are both on the same point. Often these implementations are heavy to ingress or computationally intensive. An autonomous system for tracking insects in controlled environments has been developed by Tucker Balch, Zia Khan and Manuela Veloso [2], and when deployed has been observed to be effective for tracking single agents in a multi agent colony and to be lighter an CPU load. There method is to first use a color region segmentation algorithm to discern which segments may be ants and then matching those with a standard frame differencing to identify an ant region. Because there is no stable background at initialisation an algorithm is placed to find a 'running average' of a background standard:
Bij = ( -1)Bij + Iij
where Bij is the pixel at (x,y) of the background image, Iij is the corresponding pixel in the current camera image, and is a parameter specifying the "learning rate" at which the background image adapts to changes. Once a region has been identified as an ant keeping track of it requires a form of data association, where we assume that the ant will only move within a set space of any given t and t+1. A problem arises when an ant moves further away from it's previous point where another moves closer to it. This in effect switches the ants. An improved method for tracking generates all possible sets of matching points between t and t+1 and calculate the 'fitness' of each to a criteria of least squared. where are the ith matching pair and is the Euclidean distance. The set of matched which minimises the above criteria is chosen [2].

Are Ants Automata?

Autonomous agents fill the ant paradigm (or visa versa) in that each agent has a set of built in rules and functions that are executed depending on the neighbouring entities, there position in the environment and the assigned task which they are carrying out. "An autonomous agent is a system situated in, and part of, an environment, which senses that environment, and acts on it, over time, in pursuit of its own agenda. This agenda evolves from drives (or programmed goals). The agent acts to change the environment and influences what it senses at a later time" [16]. In autonomous artistic systems the work evolves on it's own that requires little or no externalized input where the creation is done by the system.
In the case of an ant colony the individual agents operate in automata singularly and collectively making a dynamic, living system. Automated composition (a.k.a algorithmic composition) is where a process is put in place that makes for minimal input with maximum output of artistic creation, hence why autonimous agents are considered a topic. Data sets can be collected from any source that has dynamic attributes. Once the data is presented it's a question of what to do with it. Fabio Cifariello Ciardi used live data from stock market environments for real time signification in his system "sMax" [17]. Andrea Polli sonified data of atmospheric and weather conditions [18] in installation for museum. The speakers in the installation space were positioned in a representation of the location of several sites where the environmental data was collected from. Meteorological data was collected from a weather station in Garth Paine's "PLantA" [19] and augmented into a sound space. Using data from sources makes use of a message or a grounds for contradiction that roots a work inherently in it's propositions. Thoughts can then be given as to the construct of the generated product: natural or artificial, is the output comprehensible, does the output allow for cognition of location. Using ants as an input for an algorithm to interpret their movement into another metaphysical structure, the message is already created.

References:

[1] Bert Holldobler, Edward O. Wilson, "The Ants", Harvard University Press, 1990
[2] Tucker Balch, Zia Khan and Manuela Veloso, 2001, "Automatically Tracking and Analyzing the Behaviour of Live Insect Colonies" AGENTS'01, May 28 - June 1, 2001
[3] Asia Pacific Ant Farm, 1994, Takamatsu City Museum of art, Japan, Yukinori Yanagi
[4] Eight Bit Ant Farm, 2002, Remo Campopiano and Jonathan Schull
[5] David Lynch. 1986. "Blue Velvet". 45min. Dino DeLaurentiis Entertainment Group.
[6] Richard Lerman. 1996. showing at 'Sound Culture festival'. San Fransisco.
[7] Steve Peters. 2004. "Here-ings: a sonic geohistory". March 27 - April 28 2004, Dorsch Gallery, Miami, Florida.
[8] Franziska Kluegl, Frank Puppe, Ulrich Raub, Juergen Tautz. 1998. "Simulating Multiple Emergent Phenomena - Exemplified in an Ant Colony". Artificial Life IV, MIT Press, 1998
[9] Ranxiao Frances Wang and Elizabeth S.Spelke. 2002. "Human spatial representation: insights from animals". TRENDSin Cognitive Sciences Vol.6 No.9 September 2002. pg 376
[10] Marco Dorigo , Eric Bonabeau , Guy Theraulaz. 2002. "Ant algorithms and stigmergy". Future Generation Computer Systems 16 (2000) 851–871, Elsevier
[11] Rebecca Booker. 2004. “What Ants and Bees Tell Us about Social Network Theory”. AAAS Dialogue on Science, Ethics, and Religion on January 22, 2004.
[12] Frank Z. Brill, Glenn S. Wasson, Gabriel J. Ferrer, Worthy N. Martin. 1997. "The effective field of view paradigm: adding representation to a reactive system". Engineering Applications of Artificial Intellegence 11, 1998, pgs 189 201
[13] October 2000 JASA issue, Hickling and Brown
[14]Markl, H. 1968. ‘‘Communication by stridulatory signals in leaf-cutting ants. II. Production and characteristics of the signals,’’ Z. vergl. Physiol. 60, 103 – 150.
[15] Flavio Roces and Jurgen Tautz. March 2001. "Ants are Deaf". Theodor-Boveri-Institut der Universita ? t Wu ? rzburg, Lehrstuhl fu ? r Zoologie II, Biozentrum, Am Hubland, D-97074 Wu ? rzburg, Germany
[16] http://en.wikipedia.org/wiki/Autonomous_Agent. available 03/03/06.
[17] Fabio Cifariello Ciardi . 2004."sMax: A Multimedia Toolkit for Stock Market Data Sonification". Proceedings of ICAD 04-Tenth Meeting of the International Conference on Auditory Display, Sydney, Australia, July 6-9, 2004..
[18] Andrea Polli. 2004. "Atmospherics/Weather Works: A Multi-channel Storm Sonification Project". Proceedings of ICAD 04-Tenth Meeting of the International Conference on Auditory Display, Sydney, Australia, July 6-9, 2004.
[19] Garth Pain. 2004. "PLantA". ICAD Sydney Opera House. July 6-9, 2004