This LEGO MINDSTORMS robot is NOT programmed to walk. That’s nothing really new. Instead, this robot is programmed to learn how to walk on its own, without any supervision. The robot crawls using a “foot” with 2 DOF (Degrees of Freedom), and senses the world through a LEGO encoder, attached to the NXT using an adaptor cable.
How does the unsupervised learning process work? At the beginning, the robot lives its childhood, and explores randomly its action-space. Some actions will produce a reward in terms of forward movement, others will produce a punishment, when the robot goes backwards by mistake. This reward/punishment mechanism is fed by the sensorial input provided by the wheels, to which the encoder is attached. When a certain amount of time has passed, the robot becomes adult, and moves using the acquired experience, rather than random. The algorithm used is known as Q-learning. We can say that every time you start the program, a new life is born. And I dare to say that quitting the program without saving the brain of the creature before, is like killing it! I have to give credit to Dr. Frank Vanden Berghen, who wrote the Robot Java Applet that inspired me.
The Q-learning robot snail was realized for the purpose of illustrating a nice example of Artificial Intelligence in the Italian TV show 1×1, aired on 21 April 2008 on SAT 2000 channel. Here I am with the MC Arianna Ciampoli, describing the working of an early prototype of the snail. That version walked by pushing, rather than pulling, since the arm was on the back.
