German-based robot manufacturer KUKA is showing off one of its small robots’ skills as part of a robotics course at OTH Regensburg (Technical University of Applied Sciences). At the university, students are teaching the KUKA KR AGILUS robot how to play a famous board game—Settlers of Catan.
Dr. Martin Weiss and Markus Webert work with the KUKA small robot KR AGILUS. (Image Credit: KUKA Robotics)
After conducting some basic research in software engineering, the students came up with this board- game project that even helped student Markus Webert finish his bachelor’s degree in computer science.
In the video, the robot grabs an orange street and places it on a screen that simulates the playing surface of the board game.
“Before I started my bachelor’s degree I thought, ‘Programming, I don’t like that much,’ but now I’m really into it,” said Webert, on KUKA’s website. “When I started to get into programming I had a professor that suggested I take part in this really cool project to program an AI that could play Settlers of Catan. We started with an online game platform that required human players, and we used that platform to collect data from them. The goal was to use that data for stuff like Machine Learning later on, so we could teach our program how to play.”
According to Dr. Martin Weiss from OTH Regensburg and Webert’s advisor, when students work with robots in this manner, they can witness the pieces moving and the machine doing some interesting things, which gets them really excited about the field.
“They like it because they can put their ideas in very practical things. Sometimes students are afraid of mathematics, but the problems can be separated in such a way that one group works on the mathematics, the other works on programming and others work on the engineering, so I think the students get the point that this is something that requires teamwork, and that this is a job where they might have a great future ahead,” said Weiss.
The students would also like to install a camera on the KR AGILUS so that future students can develop vision systems that allow the robot to pick up pieces at random; but for now, the robot will continue to play as it knows how.