ABSTRACT
The navigation stack is one of the most popular capabilities in robot operating system (ROS). It incorporates consideration of a priori maps, dimensions of the robot, mobility of the robot, and obstacles perceived en route. This chapter begins with describing a process for map making, then shows how maps can be used with the nav-stack and how action clients can invoke the functionality of the nav-stack. The navigation stack incorporates multiple components, including static and dynamic costmaps, a global planner, and a local planner. The purpose of the local planner is to respond to unexpected obstacles encountered while following the global plan. This approach can behave well if the mobile robot does a good job of following each local plan, since iterations of re-planning would then result in reiterating the formed plan. With the robot running, rviz $ \mathtt rviz $ should show that the LIDAR is active and that it can sense the walls of the starting pen.
