An Integrated Simulation Platform to Validate Autonomous Vehicle Safety
Autonomous driving systems rely upon sensors and embedded software for localization, perception, motion planning and execution. Autonomous driving systems can only be released to the public after developers have demonstrated their ability to achieve extremely high levels of safety. Today’s hands-off autonomous driving systems are largely built with deep learning algorithms that can be trained to make the right decision for nearly every driving situation. These systems, however, lack the detailed requirements and architecture that have been used up to now to validate safety-critical software, such as that which controls commercial airliners. Road testing is clearly an essential part of the development process, but billions of miles of road testing would be required to validate the safety of autonomous driving systems and software. Simulation is needed to make verification and validation of the operation of autonomous vehicles a practical endeavor.