Research

In WP-A we will investigate how the emergence of AVs and connectivity might change the way human drivers behave in traffic.

Traffic flow efficiency and safety are the consequences and result of the interactions between vehicles. We lack proper understanding of how these interactions will change when AVs are introduced, and how human drivers adapt their behaviour when interacting with AVs on such demanding road sections. Behavioural adaptation, is an important human factor that will affect the dynamics of mixed traffic. The two main research objectives of WP-A are:

• RO-A1: To understand human drivers’ behavioural adaptation when interacting with AVs, and to develop a behavioural theory and mathematical models for these interactions;
• RO-A2: To investigate the implications of CAVs penetration rate on drivers’ behavioural adaptation.

In WP-B we will develop algorithms and models which expands AVs Operational Design Domain (ODD).

To deal with infrastructure peculiarities and complex traffic interactions safely and efficiently we need to increase AVs capabilities by understanding how to expand the ODD. The interactions in mixed traffic are dependent on AVs’ capabilities and limitations, i.e. the ODD. Among the main determinants of the ODD are two types of interactions: first, the interaction of AVs with the infrastructure; and second, the interactions of AVs with other vehicles. The two main research objectives of WP-B are:

• RO-B1: To develop a methodology for peculiarities identification on different roads and traffic conditions, and to develop accurate and reliable algorithms for peculiarities features’ extraction, recognition and prediction using data driven approach;
• RO-B2: To examine and evaluate the implications of different driving strategies and driving styles of AVs on human-drivers’ behaviour of nearby vehicles.

In WP-C we will implement the new developed behavioural models in WP-A and WP-B in existing open source simulation platform and assess the implications of mixed traffic on traffic flow efficiency and safety.

Previous studies have already used microscopic simulation tools to assess the effect of the longitudinal control task of automation, i.e. ACC (Adaptive Cruise Control) and cooperative ACC (CACC) systems with V2V communications on traffic flow efficiency and stability, as well as in mixed traffic. However, these studies reached widely varying results because of different assumptions about the behaviour of human drivers and automated systems. While using simulation is a reasonable compromise in this circumstance, there is a high risk of oversimplification because an important component, human behaviour adaptation when interacting with AVs, is not accounted for. Therefore, we will investigate the importance of this assumption and its impact on the simulation results. This is a prerequisite to have a reliable simulation tool for mixed traffic. The two main research objectives of WP-C are:

• RO-C1: Implementing the new knowledge on humans’ behavioural adaptation when interacting with AVs, and AVs behavioural models in an existing open source simulation platform;
• RO-C2: Assessing the implications of different scenarios on traffic flow efficiency and safety, and consequently propose recommendations regarding the infrastructure (physical and digital) requirements.