Students’ robots power through maze in Artificial Intelligence course
May 12, 2011 By Leigh Cocanougher
created “wall follower” robots that navigated a floor maze.
“Implementing something that works,” Shannon says, “is always more
satisfying than just studying the theory.”
On Monday, May 9, an unusual activity took place in the basement of Boles Hall. There, finding their way around a large maze on the floor, were “wall followers,” robots programmed by students to stay near the maze walls and follow its shape.
The project was an assignment in mathematics and computer science professor Christine Shannon’s “Artificial Intelligence” course, a class with an emphasis on how to create intelligent agents “in the sense that they can make decisions based on their environment,” Shannon says.
“Basically,” she explains, “an agent is supposed to have sensors to take in information from the environment and then act based on that information and whatever other information it has at its disposal. These robots give us an opportunity to create solutions to small problems that can be completed in a week or 10 days but still illustrate main ideas. Implementing something that works is always more satisfying than just studying the theory.”
To enable the robots to navigate the maze, the students equipped them with touch sensors (which notify the robot that it has bumped into something), light sensors (which allow them to make decisions based on the amount of reflected light) and ultrasonic sensors (which allow them to approximate the distance to an object).
In addition, students were asked to add a unique and imaginative function to their creations. “One group collected the information about turns and distances traveled while it searched for a black object and then wrote them to a file,” Shannon says. “If this were a search robot prowling through some wreckage, it could relay the directions to a victim.”
Another robot shot off a projectile when it found the black square, and another played a song when it reached its goal. One group’s creation even mounted the ultrasonic sensor on a rotating shaft to allow it to look in several directions.
Shannon hopes that the project—and the course as a whole—enhanced her students’ programming skills and enlarged their view of the ways in which computing can be applied.
“I also hope that they’ve gained a renewed appreciation for what humans can do,” she says. “Many things we take for granted—like our ability to use language—are extremely difficult to duplicate in a meaningful way.”