With recent introduction of affordable sensor and actuator hardware, as well as a huge increase in computing power, the wide adoption of intelligent robots is finally becoming a reality. Such a movement is greatly accelerated by recent robotic competitions, which requires a general, self-contained robot system that can autonomously work in uncontrolled environments. In this talk, we will show how we build a modular, general purpose robotic software framework from simple and efficient software modules, and how the software framework has been successfully used for two recent international robotic competitions, the RoboCup international robotic soccer competition and the DARPA Robotics Challenge.

Seung-Joon Yi is an assistant professor in the Department of Electrical Engineering at the Pusan National University. He received his B.S. in electrical engineering and Ph.D. in machine learning from Seoul National University, and worked at the University of Pennsylvania as a visiting scholar and a postdoctoral researcher. Before joining PNU, he was a researcher at Naver Labs. His main research interest is leveraging machine learning approaches to build more intelligent robotic systems. He led the software development for the UPENN robotic soccer team, which earned the world championship for five years in a row, as well as the UPENN - UCLA joint team for DARPA Robotics Challenge Trials and Finals. He received a silver medal in the 35th International Mathematical Olympiad (IMO) held in Hong Kong, 1994, and he was awarded the URAI Young Scientist Award in 2014.