Teaching Autonomous Navigation Using an Open-Source Middleware in a Hybrid Format

Along with other topics in mechatronics, autonomous navigation has been advanced amazingly in recent decades and is now playing a pivotal role in many industrial automation applications.This paper discusses the new strategies designed to teach autonomous navigation of mobile robots using robot operating system (ROS), the most popular open-source robotics middleware.At Michigan Tech, the authors offered the course to undergraduate seniors and graduate students in a hybrid format with live broadcasted recitations and in-person labs.As the time and capacity of the lab were limited, the instructors put their effort into providing resources that students could work on implementations on their own.The main difference from the past teaching was providing Gazebo simulation environments and template codes with ROS.Based on the recitations taught before the lab sessions, the students are asked to write the core of the algorithms in Python using the provided templates.Providing simulation tools helped the students implement and debug their algorithms before testing their codes on the actual robots.While students could practice and experience autonomous navigation on their own through the simulations, they could experience the difference between the simulator and the real-world robot motions.Many industries already use ROS for their implementation in the field, and many commercial robots are available for ROS.Therefore, the authors believe this newly designed course provides a more profound knowledge and practical experience in mechatronics.