Robotics simulation – A comparison of two state-of-the-art solutions

摘要

Simulation plays a significant role in development and deployment of robotics systems. Depending on the planned tasks, different functionalities shift into focus. For mobile robotics systems using the Robot Operating System (ROS), Gazebo has been the goto solution for simulating behavior, sensors, and interaction with environments for many years. With the availability of numerous novel solutions, a comparison of the different strengths from a use-case perspective becomes necessary. In our paper, we compare Gazebo and NVIDIAs Isaac Sim in the context of ROS-based autonomous mobile robots. Special attention is given to the integration of the simulators in the ROS-environment and their ability in regards to sensor simulation. This paper aims to provide a comparison between different features in Gazebo and NVIDIA Isaac Sim for simulating mobile robots and sensor systems.