Comparison of Simulators based on Scenario-based Testing of Autonomous Vehicles

The future of modern transportation could be greatly improved by autonomous driving. However, its reliability and safety have drawn a lot of attention and concerns. This new paradigm poses unique challenges for testing. The expense and risk of testing an autonomous car algorithms on actual an autonomous vehicles are prohibitive. So many researchers cannot afford real AVs for testing on the roads. However, comprehensive and rigorous testing is crucial for ensuring that AVs can manage the wide range of conditions that can be seen on public roads. It is expensive and risky to test AVs on public roads. This is a challenge that is addressed by simulation testing, However, the disadvantage of simulation testing is that it is only as effective as the testing simulator and the degree to which the simulated situations faithfully represent the actual world. The fundamental characteristics that a successful simulator should have are outlined in this article. Further, we provide a comparison of commonly used simulators namely CARMAKER, CARLA, and LGSVL. Our analysis reveals that, for the reasons stated in this work, the Carmaker simulator is superior to the most cutting-edge simulators available now for testing self-driving cars end-to-end are CARLA and LGSVL. Analysis of simulator is based on Pedestrian scenario testing in different software. We demonstrate that by using this software, we can more efficiently test scenarios for an autonomous vehicles.