Autonomous Ground Navigation in Highly Constrained Spaces: Lessons learned from The BARN Challenge at ICRA 2022

To compete in The BARN Challenge, Designing autonomous robot navigation systems has been each participating team needed to develop an entire software a topic of interest to the robotics community for decades [1]- stack for navigation for a standardized and provided mobile [5]. Indeed, there currently exist many such systems that robot. In particular, the competition provided a Clearpath allow robots to move from one point to another in a collisionfree Jackal [26] with a 2D 270 -field-of-view Hokuyo LiDAR manner (e.g., open-source implementations in the Robot for perception and a differential drive system with 2m/s Operating System (ROS) [4]-[6] with extensions to different maximum speed for actuation. The aim of each team was to vehicle types [7]), which may create the perception that develop navigation software stack needed to autonomously autonomous ground navigation is a solved problem. This drive the robot from a given starting location through a dense perception may be reinforced by the fact that many mobile obstacle filed and to a given goal, and to accomplish this task robot researchers have moved on to orthogonal navigation without any collisions with obstacles or any human interventions.