The present doctoral dissertation attempts to investigate the introduction of automation in road transport systems by focusing on impacts of autonomous vehicles on traffic and by taking into account all possible challenges. More specifically, the aim of the dissertation is to define driving behavior profiles that simulate the behavior of autonomous vehicles and cover all relevant levels of automation for different market penetration rates of connected and autonomous vehicles. In this context, the profiles will be investigated through a microscopic simulation process, while their identification will be carried out through the development of machine learning models. The results could enable drivers and stakeholders to be prepared about the introduction of automation in urban environments and could be also beneficial for future management of cities.