Teach Reinforcement Learning in Robotics Class: two case studies
Yonghe Zhang, Wenjiao Qu, Yundan Xiao, Guocheng Zhong · 2022
Reinforcement learning (RL), as a promising approach to realize Artificial Intelligence (AI), has been widely applied in developing robots. However, the current K12 Robotics Course is still designed to focus on traditional robot technology without RL. In order to foster the transition between education and industries, a robot course integrating RL is designed in our research. We implemented the course and carried out case studies in a robot course in a school and an experience class for K12 students. The results show that most students can successfully complete the robot task based on RL under reasonable course difficulty, and most students agree the proposal of teaching RL in K12 Robotics Course. This suggests researchers and policy makers to work on updating K12 Robotics Course by adding RL section, thus to keep the AI curriculum system and content modern.