Design of a Multitask and Terrestrial Robot Challenge for Course Project of Intelligent Robots

Abstract

Robotics is a highly practical, multidisciplinary, and rapidly developing field of study which lies at the intersection of computer science, automatic control, mechanical engineering, and materials science, as well as other disciplines. Designing a suitable yet challenging robotics course project is incredibly important and beneficial for teaching the subject effectively. Carrying out a practical robotics project helps students consolidate their knowledge and connect concepts across disparate areas of computer science and mechatronics engineering as they design and develop an integrated robotic system. Additionally, a practical project helps deepen a student’s understanding of how recent ground-breaking results in Artificial Intelligence (AI) can be implemented in a real robotic system. This paper provides an overview of a project for the Intelligent Robots graduate course. In the proposed course project, each team programs a small-scale humanoid robot to carry out a multi-task robot challenge on a terrestrial map. The course project is organized into a series of milestones, each of which is related to the corresponding module of the robot (e.g., joint motion, visual servoing, etc.). The students’ performance in the course project over the past three years has been summarized and reviewed. The evaluation results and the student feedback show that the proposed course project is interesting and engaging, and provides the students with a comprehensive opportunity to apply their theoretical knowledge in robotics, as well as to implement the findings of contemporary research in AI, to a real robotic system.