부산대학교 도서관

This paper proposes methods to compare two different navigation frameworks: move base—Potential Gap and Global Path Follower—Potential Gap in outdoor environments. Initially, we plan to utilize Gazebo and RViz simulators to first run test scripts in a simulated map. To assess outdoor navigation performances, we decide to use success rate, navigation time, trajectory length, and velocity smoothness as our evaluation metrics. Then we will move into real-world experiments. For outdoor testing, we decide to run trials on an open space platform and a narrow alleyway and plan to implementing Monte Carlo Localization to pinpoint the location of TurtleBot by combining GPS and robot odometry data. After completing all the trials, the final results will be analyzed and included in future works.