VRU and CAV interactions


The overall aim of this PhD research program is to prepare Queensland for a safe integration of Connected and Automated Vehicles (CAV) into mixed urban traffic environment by observing and predicting CAVs and other road user interactions.

The key focus of the study is to investigate Vulnerable Road User (VRU) receptivity, mental models (e.g. communication strategies, situation awareness and trust), and behavioural adaptations towards road sharing with CAVs.


Project background

Connected and automated vehicles (CAVs) as an emerging innovative technology is deemed to have great potential in improving road safety, mobility and environment. Based on the six levels (0‐5) of automation defined by SAE International, highly automated driving systems (SAE Level 4) in certain environments and certain conditions can perform all aspects of driving tasks, and a human driver is not required to control the vehicle in any circumstances.

The introduction of highly automated vehicle signifies that the traditional on‐road interaction mode between drivers and other road users (e.g. pedestrians, cyclists, motorcyclists) such as eye contact or waving a hand would not occur. Therefore, even though the automation system has great potential in the future, it remains unknown as to what extent the interaction with other road users is safe, and how conflicts could be prevented.

As CAVs are likely to be deployed initially in mixed traffic, they need to interact safely and efficiently with other road users, including manually-driven vehicles, motorcyclists, e‐scooters, cyclists, and pedestrians.

Therefore, there is a need to understand the pattern of physical interactions between CAVs, motorists and VRUs to assess potential crash risks. How VRUs will respond to the presence of CAVs when they are sharing the road (e.g. at intersections) and how CAVs should react when they detect a particular VRU’s behaviour are fundamental research questions required in order to understand the safety benefits of CAVs.

An interaction model could be expressed as a simple gap acceptance measurement. A gap acceptance criterion could be based on parameters derived from probabilistic distribution to take into account the heterogeneity of the VRU population. A more comprehensive approach is to model the interaction with intelligent agent‐based simulations to investigate and predict the emergence of complex group behaviour through simulating the actions and interactions of a large group of autonomous agents in given driving scenarios (e.g. urban traffic).

The resulting interaction model will benefit road safety stakeholders as it will be used for assessing future VRU / CAV crash risks. The project will model the interaction between VRUs and CAVs to:

  • inform the design of road infrastructure of the best way to minimise road crashes
  • guide the development of CAV safety test scenarios for ANCAP (including virtual simulations)
  • provide interaction model for future traffic or agent‐based simulation (e.g. AIMSUN)
  • provide educational recommendation to road safety stakeholders including policy makers on how VRUs and CAVs should share the road safely

Project objectives

The general aim of the study is to develop a CAV‐VRU interaction model using advanced mathematical and computational techniques. The model will highlight potential conflict or crash risks.

The model will be tested experimentally using the Queensland Department of Transport and Main Roads’ ZOE2 vehicle, and will incorporate a number of human factors and Queensland-specific factors. The specific objectives include:

  • defining formal (mathematical) models of interaction between CAVs and VRUs
  • design novel methods for assessing road user intentions and predicting cooperative behaviour of road users
  • establish crash risk assessment of interactions between CAVs and VRUs
  • investigate Queensland‐specific factors affecting the interactions/cooperation between CAV and VRUs
  • inform road authorities, public authorities, urban planners about crash risks associated with particular road configuration / geometry
  • formulate policy recommendations on regulation of interactions between VRUs and CAVs

More from iMOVE CRC

Most Popular

To Top