About

Current benchmarks for dynamic obstacle avoidance do not provide a way to alter how obstacles move and instead use only a single method to uniquely determine the movement of obstacles, e.g., constant velocity, the social force model, or Optimal Reciprocal Collision Avoidance (ORCA).

We introduce DynaBARN, a simulation testbed to evaluate a robot navigation system's ability to navigate in environments with obstacles with different motion profiles, which are systematically generated by a set of difficulty metrics.

DynaBARN contains:

• 60 Gazebo environments populated with dynamic obstacles of varying trajectories
• Metrics to quantify the difficulty of the environments
• Benchmarks for DWA, TEB, Reinforcement Learning, and Behavior Cloning planners
• Code to crete more environments to the user's liking

DynaBARN is customizable for your robot's specific size, while we provide 60 pre-generated environments for small-sized Unmanned Ground Vehicles, e.g. a ClearPath Jackal robot. We also provide a set of difficulty metrics to test your navigation system's sensitivity to different navigation difficulty levels.

Example Worlds

Here are examples of worlds of different difficulty with a reinforcement learning policy deployed in it:

Links

• DynaBARN Paper
• Pre-Created 60 Worlds
• DynaBARN Code

Contact

For questions, please contact:

Dr. Xuesu Xiao
Department of Computer Science
The University of Texas at Austin
2317 Speedway, Austin, Texas 78712-1757 USA
+1 (512) 471-9765
xiao@cs.utexas.edu