AO2000

Project Overview

The Automation Orchestra 2000 is an interactive musical robot made up of various ordinary household appliances, motors, home-made electronics, an acoustic guitar, and a cast-off computer. The music is produced by the motors striking different surfaces, and the appliances cycling on and off. AO2000 interprets audience location information, which changes the automation sequence of the music. The crucial aesthetic of the piece’s construction is its intentional use of low-tech materials, and its kludge-style execution.

Goals

My proposal to the Berwick Research Institute (BRI) was to build a robot for the Artbots 2002 competition. The first of its kind, the competion was an exhibition of creative robots, showcasing the artistic potential of emerging technologies.

One of my goals for the project was to create a metaphor of the deficiencies in language. I wanted to investigate the gaps or bridges of communication and of the multiplicity of meaning. Another goal was to use what was available, to cobble together an overly sophisticated system from simple components, rather than ordering brand-new, highly specialized equipment. During my residency, I conducted multiple tests that eventually became the building blocks for AO2000.

Technical Description

The robot’s primary brain is a 286 laptop computer I bought at the MIT Flea Market for $5. A program written in QBASIC interprets keyboard input and controls the sequencing of the music. Closed-circuit cameras observe the movement of the audience around AO2000. Light-activated relay boards attached to the TV screens trigger solenoids to pound the keyboard. These mechanical keystrokes tell the computer to automate different instruments. The computer controls the instruments via the printer port, which controls an 8-channel relay board. The relays turn the power on and off to the two blenders, two egg beaters, four groups of motors. The motors’ spinning shafts are attached to lever arms and stiff cords, striking and resonating the guitar’s strings, and in general causing a loud clattering.

Aesthetic Description

I knew that my competitors from universities, would have better resources than I would. Having access to those resources, my piece could have been realized using more sophisticated technical means. I could have used Max/MSP and NATO to interpret the audience’s movements. I could have used pre-fab modules designed for industrial automation. I could have designed the robot with a 3-d CAD program and had the parts laser-fabricated from aircraft-grade aluminum.

I knew I couldn’t compete along strictly technological grounds, coming from a non-profit arts lab without a robotics background.  My solution was to go in the opposite direction. I deliberately built my robot as low-tech as possible, making clever use of materials that I had at hand. This methodology became a commentary on the techno-fetish trend in contemporary electronic arts. In addition, the robot processed information in an intentionally convoluted way. It would have been more direct for the light sensors to connect directly to the computer via an input port. Instead, the system of closed-circuit tv, light sensors, and solenoids formed an individual robotic sub-system, which communicated with the analysis and controlled the robot through an interface designed for humans; a keyboard. By making the interface needlessly convoluted, AO2000 illustrated deficiencies in communication and the comprehension of complex meaning.

Results

During my residency, the project became part of a larger experiment to see how I could make this thing actually work and sound good. My strategy of building upon a multitude of small tests was successful, because many of the experiments failed. What came out of this process was a series of modular building blocks that could be adapted to different situations.

AO2000 was a success in that it worked consistently without glitches, and that it won the Artbots competition. Its biggest drawback was that it didn’t have enough sonic range or diversity. On average, the result was a cacophony of sound.

It also became a model of how to work, by developing projects incrementally through trial and error. AO2000 is one project in an ongoing series of investigations. The newest incarnation is a quieter model that attempts to resolve the problems of its predecessor, while reusing some of the same components.