UM News

The RoboCanes who competed in Nagoya, Japan are, from left, Chloe Arluck, Pedro Peña, Lloyd Beaufils, Professor Ubbo Visser, Joe Masterjohn, Jannes Visser, Kyle Poore, and Andreas Seekircher.

Despite a year hiatus and outdated hardware, the RoboCanes, UM’s autonomous soccer playing robot team, advanced to the semi-final round of RoboCup 2017, becoming the world’s fourth top team in the annual robotics and artificial intelligence (AI) competition.

Their secret? A new strategy, according to Ubbo Visser, associate professor of computer science, who guided his five-member team of computer science students to the games held in Nagoya, Japan, late last month.

“After one year without participating in the 3D Simulation League, we were able to submit a competitive package and compete among the 12 teams and reach the semi-final, where we lost against the long-time world champion, UT Austin,” Visser said.

Designed to promote robotics and AI research, the international robocup challenge is as much about learning as it is about fun. To compete, students must apply integrated research to develop a myriad of robotic skills, such as real-time sensor fusion, reactive behavior, strategy acquisition, machine learning, real-time planning, multi-agent systems, context recognition, real-time computer vision, strategic decision-making, motor control, intelligent robot control, and much more.