Inspired by the book Vehicles, Ana BBot is a sophisticated Braitenberg robot -- a robot controlled by an analog neural network. On board electronics implement a collection of simple McCullough-Pitts neurons that can be configured to give the "Braitenbot" a number of simple behaviors. These include "fear" (obstacle avoidance), "love" (object attraction), "searching", and "schooling". Braitenberg vehicles 1 through 6 can be implemented on this robot.
In terms of hardware, Ana has two analog range sensors and two analog light intensity sensors. These can be connected, optionally through up to six neurons, to two analog wheel motors. All "programming" is done through jumper wires which connect analog signals from some source to another destination. No separate computer or downloading is necessary -- configuring the neural network is deliciously hands-on and intuitive. Moreover, you can watch the input and output signals vary at each level of neural circuitry using the built-in meters. This provides a uniquely direct insight into the functioning of the creature.
For some more detailed information, check out this instruction sheet.
Programmable with jumper wires!
Runs on 4 AA cells!
Sensors can be moved around!
Contact the folks at Johuco (firstname.lastname@example.org) if you have any questions.
Return to Johuco's home page
Last updated 8/24/06.