Implementation of SAE Automation Levels 2 and 3 on a Driving Simulator for Behavioural and Road Safety Studies

The transportation industry is experiencing a significant transformation due to the advancement of vehicle automation technologies. This transformation is marked by various levels of automation that represent major milestones along the path towards achieving fully autonomous driving. SAE International presents a classification scheme with detailed definitions for six levels of driving automation. Briefly, the levels of driving automation are determined by the extent to which the responsibility for driving shifts from the driver to the vehicle. While the shift in driving responsibilities introduces new paradigms and practices, existing road infrastructures and current road design guidelines have predominantly considered the human behaviour on manual driven vehicles. This project was conducted through a collaborative partnership between the TRAFIKKLAB at Nord University, Norway, and the Road Safety and Driving Simulation (RSDS) Laboratory at Politecnico di Torino, Italy. The principal objective of the project was to establish a foundation and develop vehicle models within a driving simulator that operate at specific levels of automation i.e., Level 2 and Level 3, enabling the conduction of simulation experiments. Research in this area help to bridge the gap between current manual driving practices and appearing automated driving paradigms by updating and improving the rules and regulations governing our roads. An ongoing research project called DIRECTIONS at the Politecnico di Torino (PoliTO) aims at conducting a series of simulation experiments focusing on the effects of road geometric features on driver behaviour and safety with partially automated driving (SAE Level 2, PAD) and conditionally automated driving (SAE Level 3, CAD). The project faces a challenge in that SCANeR Studio® simulation software does not have predefined vehicle models for these intermediate levels of automation. Consequently, the project's first objective and the main scope of this thesis is to develop level 2 (PAD) and 3 (CAD) automated driving modes, tailored to the needs of the DIRECTIONS project. Instead of developing a new automated driving system (ADS), a straightforward strategy has been chosen that makes use of native functionalities of the driving simulation software to replicate the level 2 and level 3 automated vehicle behaviours under carefully defined experimental conditions. The project is supplemented by the development of a human-machine interface (HMI) for each level of automation designed to convey essential information to the driver and by incorporation of mechanisms for smooth driving mode transitions between human driver and automation system. Also, several sensitivity analyses were performed to gain insights into the diverse range of software settings and parameters, in order to offer guidance on customizing the models to suit the precise demands of the simulation experiments. Finally, after thorough testing and evaluation, it was evident that the models performed effectively within the simulator environment, the Human-Machine Interface (HMI) consistently and promptly conveyed essential information to the driver, the model were capable to accurately collect the necessary data, and the implemented mode transition mechanisms operated as intended, ensuring safe and smooth transitions between human and automated driving modes.