In this talk you will learn how to model high-fidelity sensors for robotics simulation in ROS (1 & 2) applications using Unity. You will learn about a variety of sensor types with different noise models, how they communicate with ROS, and how to debug them using Unity’s visualization framework. We will walk through the simulation of an end-to-end, closed-loop control scenario that utilizes several sensor types, including an IMU, LiDAR, RGB camera, and RGB-D camera. We will also compare algorithms and ROS packages for object detection, localization, and state estimation using the sensory data under various settings and noise models.
