Development of an Unity-ROS2 Simulation Environment for SLAM evaluation on autonomous vehicles

Kurzfassung

This project aims to develop a realistic simulation environment using Unity and ROS2 for evaluating SLAM algorithms in autonomous vehicles. The simulation incorporates sensors such as LiDAR and IMU, as well as adverse environmental conditions, like rain, to assess the robustness of perception and navigation systems. Specific scenarios emulating complex and dynamic environments, such as urban canals, were designed to enable comprehensive testing of autonomous systems in controlled conditions. The results include a rain simulation model that introduces noise and attenuation in LiDAR measurements, enhancing the validation of perception algorithms. This simulation framework not only enables the evaluation of the accuracy and reliability of SLAM systems but also provides synthetic data to improve these algorithms, contributing to the development of more robust autonomous systems prepared for real-world environments