Experiments with Reactive Robotic Sound
Sculptures
8th International Conference on Artificial Life
by Sam Woolf, Interact Lab, Dept. of Cognitive and
Computing Sciences, Sussex University
Abstract
This paper describes two robotic sound sculptures
produced in collaboration with Danish artist Tine Bech.
‘Echidna’ and ‘Boundless in
Space’. The sculptures were exhibited at the
Aarhus Kunstbygning gallery in Aarhus, Denmark from the
3rd to the 25th August 2002. I will describe the
sculptures and also defend the use of simple reactive
robotics in interactive art.
Introduction
After exhibiting my interactive sound installation
“The Sound Gallery”(see [1]) in Brighton,
England, I was approached by artist Tine Bech who was
excited by the idea of using robotics and artificial
life techniques to add interactivity to her sculptures.
Tine explained to me how after having studied large
scale land art pieces such as the Rodden Crater project
by James Turrell [2] she had come to realize that art
is something that should be experienced not only
visually and intellectually, but with the whole body
and with all the senses. As Merleu-Ponty wrote,
“Our understandings of space emerge from action,
indeed space is to be defined as a ‘certain
possession of the world by my body’”[3].
The massive scale of land-art pieces in relation to the
body force the perceiver to experience the work with
the whole body through the medium of action. And
through this means of interaction, the perceiver gains
a greater sense and understanding of not only the given
environment, but also of the nature of his own body and
its sensory apparatus too.
Robotic and artificial life sculptures encourage people
to interact in a more somatic and tactile way without
resorting to the extremities of scale of the land
artists. By building motion and proximity sensors and
sound and motion actuators into the sculptures we were
able to make the artworks respond to bodily
interactions with sound and movement. This amounts to
an invitation to play with the sculptures in a physical
mode that goes beyond a purely intellectualised
appreciation.
Description of the pieces
The two pieces that we produced for the exhibition were
entitled ‘Boundless in Space’ and
‘Echidna’. Both pieces consist of a
sculptural form built by Tine augmented with some
simple electronics built by me. Boundless in Space,
with its textured organic surface made from uncooked
tapioca pearls, looks like a creature that could have
emerged from some dark uncharted sea. Its surface is
studded with small semi-spherical growths that could
conceivably be interpreted as the wart like
manifestations of some alien disease, as an excessive
multitude of non-mammalian mammilla, as parasites or as
progenic fungal spores. Whatever they may be, several
are not attached to the main body of the sculpture and
take their chances alone, isolated on the gallery floor
and at the mercy of blundering feet.
Eight ultrasonic proximity sensors are mounted around
the sides of the sculpture, which is powered by an
electric motor from a car windscreen wiper. If
over-zealous spectators get too close it will attempt
to scuttle away whilst its multiple relay switches
click maniacally.
Boundless in Space is also equipped with a small radio
transmitter that allows it to communicate with the
trail of independent growths, some of which may vibrate
in sympathy or anger when its personal space is
invaded.
The control circuitry for Boundless in Space is very
simple, having more in common with simple Braitenberg
Vehicle [4] style robots than with more complex
artificial life paradigms. Logic gates are used to
determine the direction in which the motor should be
activated according to which sensors are stimulated.
Yet despite the simplicity of its control circuitry,
Boundless appears to display complex autonomous
behavior. If approached by an observer it will attempt
to withdraw, as if trying to flee from a perceived
predator. If approached by several people from more
than one side, Boundless jitters indecisively, as if
unsure of which way to turn. These movements are
accompanied by sharp rhythmic clicking sounds that in
fact originate from the multiple relays switches inside
the robot, but that resemble panicked warning signals
or pleas for help.
‘Echidna’ is another name for the Spiny
Anteater, a porcupine like creature that may be able to
sense electromagnetic fields [5]. Echidna’s body
is made from a tangle of wire. It sits on top of a
small plinth inside of which are hidden the
electronics, a very simple circuit based around a
phased-locked loop chip. Echidna’s entire body
acts as an antenna that picks up variations in the
electromagnetic field around the sculpture caused by
the proximity of people. These variations are converted
to sound, giving Echidna a voice.
Despite its simplicity, Echidna exhibited a large range
of interesting sonic behaviours. This behaviour
reflects not the sophistication of the underlying
electronics, but the complexity of the environment in
which the sculpture is situated.
Low tech versus high tech
The artist Ken Rinaldo is one of the most well known
exponents of A-life influenced robotic sound art. In
one paper he imagines an ideal of this kind of robotic
art as “a cybernetic ballet of experience, with
the computer/machine and viewer/participant involved in
a grand dance of one sensing and responding to the
other...[allowing] new sculptural and virtual
algorithmic manifestations which will far surpass our
wildest imaginations”[6]. It is assumed that such
a scenario would entail very sophisticated robotics and
artificial intelligence programming. In a forthcoming
paper, Bird et al argue that only ‘epistemically
autonomous’ robots capable of dynamically
creating new mappings between sensors and actuators and
of constructing their own sensors could fulfill
Rinaldo’s vision. [7] In fact we found that using
only simple electronics we were able to create
interactive artworks that inspired a great deal of
curiosity and interest through their ability to respond
with a surprising diversity of sound and movement to
intuitive interaction.
Rinaldo’s writes that “Artificial Life
techniques…present opportunities for both
artists and viewer/participants to develop true
relationships with the computer that go beyond the
hackneyed replicable paths of "interactivity" which
have thus far been presented by the arts
community.” He further claims that “The
word interactivity is often overplayed perhaps because
it demonstrates a kind of technical seduction although
there may not be a deeper aesthetic realization. In
other words, finding a real poetry of interactive form
and content is reliant upon simple button
activation”[7]. Is this an argument against the
effectiveness of simple reactive robotic systems in
art? Rinaldo seems to be implying that simple robotic
systems are only interactive in a trivial and
repetitive way. But reactive robots often behave in
ways that belie their apparent simplicity. Edward
Ihnatowicz’s ‘Senster’ robot, created
in
1969, was controlled by a 12-bit Honeywell computer
with only 4K of memory. However, Paul Brown tells us
that “Behavioural scientist queued up to do
experiments with the system and couldn’t believe
that something so simple could produce behaviour so
lifelike”[8].
In cases like this, the complexity often arises not
from the machine, but from the environment in which it
is situated. The first requirement for more than
‘button interactivity’ is environmental
sensors sensitive enough to pick up and respond to the
complexity of the environment it is in. A button is a
very impoverished input device, allowing a machine to
respond to only two external system states, i.e.: on
and off. By using analog sensors a much greater
diversity of behaviour is made immediately possible.
The simple electrometer/theremin-like sound sculpture I
built with Tine exhibited complex sonic behaviour in
the gallery. Tickling it seemed to have an entirely
different effect from pushing on it firmly and
sometimes invisible spikes in the electromagnetic field
would cause unpredictable yelps of different pitches
Boundless in Space’s jittery, nervous behaviour
is also partly attributable to the
‘noisiness’ of its cheap ultrasonic
proximity sensors. A push button is always on when you
press it and off when you don’t. Less reliable
sensors lead to a less predictable response.
It is interesting to note that people, especially
children, immediately knew how to play with the
sculptures and took to them with intuitive delight.
This is not always the case with more complex adaptive
pieces. Such pieces often leave people confused as they
find it hard to correlate their activity with that of
the machine and this can be a barrier to their
appreciation of the work. It may be that they find they
have a cybernetic ballet partner who doesn’t feel
like dancing. Rinaldo raises this point when he claims
that “One difficulty with some Artificial Life
artworks is that the systems may not seem to be
responsive to the changing environment, as the work
demonstrates its own internal desires. This can make
the work seem unresponsive or uncaring”[6].
Nielson recently wrote about the issue of good
usability in physical interaction (as opposed to
virtual interactions with screen based GUIs). He says:
“we’re simply replacing one form of syntax
with another that has more degrees of freedom. And more
choice for interaction design means more ways to make
things difficult for users.”[9]. It seems strange
to worry about the ‘usability’ of art. But
it seems to me that intuitiveness of interaction is
probably a pre-requisite for any true fulfillment of
Rinaldo’s vision.
To conclude, while the future may bring us
sophisticated Artificial Life artworks capable of
astonishing us with unimaginable novelty, we should not
forget that humble reactive robotic systems capable of
sensing and reflecting the complexity of their
environments have the capacity for unpredictable and
life like behaviour that encourages playful somatic
interaction.
References:
[1] Woolf, S “The Sound Gallery” ALMMA 2001
[2] http://www.rodencrater.org/
[3] Maurice Merleau-Ponty, The Phenomenology of
Perception, Routledge, London, 1962.
[4] Braitenberg, V. “Vehicles: Experiments in
Synthetic Psychology” MIT Press 1986.
[5] Augee ML and Gooden BA (1992) Evidence for
electroreception from field studies of the echidna
tachyglossus-aculeatus. Augee, M. L. (Ed.). Platypus
and Echidnas. Ii+296p. Royal Zoological Society Of New
South Wales: Mosman, New South Wales, Australia. Illus.
Maps. Paper. Isbn 0- 9599951-6-1. 0: 211-215.
[6] Rinaldo, K. Technology Recapitulates
Phylogeny:Artificial Life Art
http://www.artnode.dk/contri/rinaldo/index.html
[7]Bird et al.”towards epistemically autonomous
robots;exploiting the artistic potential of physical
systems”. Forthcoming.
[8]Brown, P 1996 “Emergent Behaviours: Towards
computational Aesthetics”
http://www.paul-brown.com/WORDS/EMERGENT.HTM
[9] Nielson, J “Making the Physical Environment
Interactive”http://www.useit.com/alertbox/20020805.html