Elizabeth G. robot project

What it does:
A visitor is attracted by the endoscopic images on the monitors and approaches the monitor tower. His or her face is then captured by a video camera. Three artificial neural networks (KOHONEN) process the video data, forming a face classification and robot control system. The sculpture is learning by seeing peoples faces and by listening to its own outputs. The tones generated by the robot are analyzed and fed back into the robot control network.

Basic robotic function:
- a big pump generates air pressure.
- adjustable voice bands oscillate in the focussed air stream.
- a robotic head modulates the resulting tone.

Robotic elements:
Actuators:

- voiceband positioning servo
- air compression servo
- air inlet valve motor
- 5 biderectional face- and lip solenoids
- 2 biderectional tounge solenoids
- 1 biderectional nose solenoids

Sensors:

- 1 microphone
- 1 piezo pick-up
- 1 pressure sensor
- positioning switches
- positioning potentiometers

Computers:

- 3 PC's (486/Pentium) [image - sound analysis and motion sequencing].
- 4 embedded controllers (80C537) [robot control].

next: about Elizabeth Gardner