Simulation-Based Reduction of Operational and Cybersecurity Risks in Autonomous Vehicles

The expected adoption of fully autonomous vehicles has raised concern about possible operational and cybersecurity risks associated with their safety. This research studies the use of a virtual autonomous vehicle in a simulation environment as a method to reduce such risks. In doing so this research contributes the process and development of a virtual model that mirrors a previously built prototype of an autonomous vehicle. It also contributes the design of the operational simulation environment for the model. Through the comparison of the physical and virtual model's machine learning training data, it is shown that the virtual model can act as a valid substitution for the physical model allowing for a greater potential for risk reduction.