Story posted June 28, 2005
It's not uncommon to bump into - or be bumped into - by one of Eric Chown's robots. They are popular and often comic denizens of the Searles Science Building, where the computer science chair and his colleague, Assistant Professor Stephen Majercik, use them to teach students advanced computing applications.
Interest in robotics has grown so steadily that a Bowdoin robotics lab recently was built that includes, among other things, soccer-playing robot dogs, called AIBOs. Bowdoin's AIBO team, Northern Bites, recently competed in the RoboCup U.S. Open 2005, which features teams developed by students at colleges and universities around the world.
It's all part of Chown's plan to attract more students to computer science, but he also has a more "subversive" reason for using robotics in the classroom: He thinks robots have as much to teach us about being human as we do them.
"The robot is actually an interesting platform on which to test cognitive theories," says Chown. "If you really want to get robots to do things that are human-like - or even different than humans - you need to look at how humans work.
"How is the architecture of the brain contrasted to the architecture of computers?" asks Chown. "Why are brains able to be so smart? How do parts of the brain fit together? When you program robots to do things, or work on other hard problems in artificial intelligence, you begin to see that a lot of intelligence isn't conscious at all: It is pattern recognition."
This summer, Chown has enlisted three student researchers to develop a high-level computer language that will allow virtually anyone to program AIBOs. These tools will allow students in his Cognitive Architecture course (fall 2005) - including non-CS students, such as psychology or neuroscience majors - to use robots to explore principles of cognitive psychology.