Self-driving vehicles are considered by many people and industries to define the future of transportation. The quest for autonomy is not only driven by safety and accessibility in urban transportation, but also in several other applications such as mining and farming. In fact the transformative role of such vehicles in transportation, urban planning, city landscape, energy, regulations, etc. has not been fully understood yet.
As the technology of autonomous vehicles continue to develop, there are several safety and regulation issues that are left to be addressed. A few research institute and organizations have started a move toward creating smart cities to develop and scrutinize the control strategies and train learning algorithms. To address the limitations, safety, and performance of this approach, we are looking into simulation-based learning algorithms that benefit from high-fidelity models and fast algorithms that can be run for different scenarios in parallel.