Accelerating Robotics Test and Evaluation with a Streamlined Simulation Process

Simulation is a popular approach for testing and evaluating robotic systems, but it is also a challenge due to the complexity of the autonomy stack (the software that runs the robot) and its interactions with the simulation software. This paper describes software tools that streamline the process of creating and running simulation models with unmanned ground vehicles that are using the ARL ground autonomy stack. Using these tools simplifies the tasks that are needed to conduct a simulation study by employing a map-based user interface, providing a checklist for selecting the topics that should be recorded, and automatically iterating over different combinations of configurations and maps. The streamlined process can accelerate the development of safe, reliable robotic systems.