Garnet Hertz: On Cockroaches, Electronic Arts and Technology Hype

0 Posted by - February 1, 2007 - Blog

Garnet Hertz builds robots controlled by giant hissing cockroaches. He is a Fullbright Scholar at the California Institute for Telecommunications and Information Technology, doctoral student at the University of California and co-founder of Dorkbot SoCal, a homey club for electronic arts nerds with chapters in cities all over the world. More importantly, he is a prairie boy from Saskatoon who can trace his interest in robotics to childhood days on the family farm recycling abandoned farm machinery into self-powered tricycles. Hertz is on leave from his research for the February 2 opening of a new show at the InterAccess Gallery in Toronto ( He spoke at Concordia University in Montreal on Wednesday, and those of us who were there got a whirlwind tour of Hertz’ creative trajectory from the wheat patch to Caltech's media labs.

Hertz came of age as a graduate student in the mid-1990s during the 'dotcom bubble', a time of hyper-hype about the potential of the internet to reorganize everything, including how we communicate and how democracies work. His early work focused on transforming intention into physical format by creating channels of communication linked to robotic control through the internet. He was able to create a website where anyone could log on and begin to manipulate the physical movements of a small robotic device on wheels, allowing it to crudely scratch the surface of a concrete floor in his lab.

Video: Cockroach Controlled Mobile Robot

When the dotcom bubble burst, Hertz’ interest began to drift. He stumbled on a news report about researchers in Tokyo (Roboroaches 1998) who had turned living cockroaches into manipulable robots—in effect, insect cyborgs – and he began to investigate the weird and ethically complex world of animal / machine hybridity.

There are, according to Hertz, three approaches to the animal / machine interface in the world of robotics. There is the cyborg approach that blurs the physiological boundary between biological and built systems, the mimetic approach which mimics biological physiology, and computational biology which applies the techniques of computer science and applied mathematics to solve problems inspired by biology. But is was remote controlled cockroaches and rats that most fascinated Hertz, despite also finding them “repulsive” and ethically problematic.

Hertz also began to hear about researchers who were using animals to control robots, but with interfaces that didn’t maim or destroy the animals. Survival Research Labs (SRL), for instance, had created large four-legged machines controlled through the movements of guinea pigs. And there was the fascinating work of Ken Rinaldo who was creating rather beautiful robotic installations driven into mechanical motion by Siamese “fighting fish” swimming around in bowls (Delicate Balance 1995; Mediated Encounters 1996; Augmented Fish Reality 2000). Hertz also discovered a group of German artists who had created small robotic vehicles that were operated remotely by hamsters on running wheels (Hamster Symbiotic Exchange of Hoarded Energy 1999). What these projects shared, aside from a way of bridging the animal/machine interface in a relatively harmless way for the animal, was the successful replacement of a microcontroller with a biological system. Hertz was hooked.

One of the big debates in the world of robotics is the nature of intelligence. Is it, some ask, of an abstract and computational nature? Thus, for instance, Garry Kasparov, the great Russian chess champion, played to a draw against Deep Blue, the most powerful computer of its generation, in 1997. Or is intelligence embodied? A more physically located amalgam of ordering principles that manage sense stimuli and motor control? The example Hertz gave to help distinguish the two kinds of intelligence was this: If we were to put wheels on Deep Blue and place it on the shoulder of a busy street, could it cross the road? Most likely not, despite its prodigious computational capabilities. Could a cockroach? Probably, despite its brain having only 1/10000th the number of brain cells as a human brain, and despite it’s inability to play chess. Embodied intelligence.

Hertz conceived and built a robotic vehicle that was controlled by the computationally challenged yet embodied intelligence of a giant Madagascar cockroach. Hertz' choice in insect collaborator was in part for its size and in part because of the Madagascar cockroach's tendency to hiss loudly when alarmed. It was a crude way for Hertz to gauge the emotional state of his biological micro-controller while at work.

The cockroach is placed into a harness that positions it on a trackball. As the roach's feet move on the trackball, the ball spins driving the robotics of the vehicle. The cockroach moves the vehicle around the gallery interacting with people, walls, chairs – – whatever the vehicle encounters. Sensors on the outside of the vehicle detect objects. Corresponding panels of bright lights facing the roach's pilot seat light up – Hertz' theory being that because roaches prefer darkness to light, they will instinctively move away from bright lights thus steering the vehicle away from objects.

Hertz made an important, if for some, obvious observation. Not all cockroaches are the same. Some, indeed, when placed in the harness hissed until they were removed – Hertz did not force these roaches into service. Others when placed in the harness did nothing, or very little. There were only a minority that actively “roamed” the galleries in their robotic cars. For people viewing the installation, many questions are raised. To what extent is the cockroach engaged in the shared space of the gallery? To what extent is the roach engaged in a virtual world versus a physical world? One of Hertz' goals was to expand the communicative sphere by translating insect interior life into a macro (mechanical) form.

Hertz' current research has him exploring, and questioning, the widely held faith in the progress perfectibility of technological innovation. He views technological change, and especially changes to communications technologies, as being in a state of constant flux and recycling rather than a state of progression. Media theorists, he explained, tend to greet each new communications medium with a euphoric expectation of communications utopia (with the expectations of the consuming public close behind). New media technologies are expected to allow us unfettered communicative abilities and to fully usurp old problematic mediums. The internet was expected to replace television and usher in an era of communications democracy, both problematic and largely unrealized assumptions. So too with cable television, with television itself, with radio, with the telegraph, etc.

Another interest of Hertz' is the role of hacker/artist in the development and uses of communications technologies. Unofficial, artistic and amateur innovators go largely unacknowledged within media theory discussions. But the role of the hacker/artist for Hertz is an essential one in a society increasingly enmeshed within an electronic membrane of tools, sensors, and machines. Few of us understand the technologies we depend on. The hacker/artist represents for Hertz a reclaiming of a knowledge with deep roots in the manifestation of power within technocratic cultures.

Hertz is also one of the co-founders of Dorkbot SoCal in Los Angelas ( Dorkbot was started in 2000 as a decentralized collection of local clubs where techno geeks, hardware hackers, circuit benders, and electronic artists can get together, share work, hang out, and exchange ideas. (Montreal's first Dorkbot meeting is February 1 (2007) at Concordia University, EV 615.)

When I asked Hertz about a noticeable lack of ethical engagement in electronic arts, he was not surprised…

[More to come…]

Leave a reply