Road crashes are a major global concern. According to the World Health Organization, approximately 1.3 million people die in road crashes every year worldwide, and road traffic injuries are currently the leading cause of death for individuals aged 5–29 years. Unprotected road users—such as pedestrians and cyclists—are disproportionately represented in these statistics. This situation underscores the urgent need for further research on road user behaviour to support the development of active safety systems and driving automation technologies that can help reduce crash risk. 

This summer school prepares students to understand current safety challenges in transportation and the potential of emerging technologies to address them by analysing, modelling, and predicting road user behaviour. 

We begin with an introduction to the fundamentals of traffic safety. Students will examine crash statistics and explore how tools such as Heinrich’s Triangle and the Haddon Matrix can provide insight into crash causation. We will then shift from crash data to naturalistic and driving simulator data to investigate which aspects of road user behaviour can be modelled. Two hands on exercises—one based on simulator data and one based on naturalistic data—will allow participants to experience how driving performance can be assessed in real-time systems, and how crash causation and system design can be informed by odds ratio analyses. Particular attention will be given to glance behaviour and its relevance to current Euro NCAP requirements and driver monitoring needs in vehicle automation. 

Various approaches to modelling road user behaviour will be presented from a theoretical perspective, complemented by real examples from our research at Chalmers and from industrial projects. Guest speakers from Volvo Cars and Waymo will provide insight into the rationale and current developments in behavioural modelling within industry. We will then discuss how behavioural models can support the design and evaluation of active safety solutions and automated driving systems. 

Given the growing safety concerns associated with micromobility, the programme includes a dedicated lecture on micromobility users, accompanied by a test riding activity that will give participants firsthand experience of the challenges these vehicles introduce into the transport system. 

Throughout the summer school, students will work in groups on a case based network activity, where they will discuss and reflect on the analysis and modelling of road user behaviour within a concrete user case. Each group will receive a hypothetical use case focused on a specific active safety system. They will analyze several behavioral modeling perspectives related to their assigned use case and present their findings to the full class on the final day of the summer school.  

To foster interaction and build community, four social activities are planned. On Sunday afternoon we will meet informally to get to know one another before the course begins. On Tuesday we will go bowling together. On Wednesday we will visit the World of Volvo center (https://www.worldofvolvo.com/), and on Thursday we will conclude with a social dinner. 

This year, the summer school also introduces an “AI twist,” featuring a new lecture on AI tools on Wednesday and an optional AI mentoring session on Friday.

With the aim to enhance knowledge on this topic, on successful completion of this school, graduates should be able to: 

• Explain the importance of analysing and modelling road-user behaviour to improve road safety. 

• Compute relevant quantitative and qualitative metrics in order to analyse and model road-user behaviour. 

• Illustrate different approaches to modelling road-user behaviour for system development. 

• Identify and experience the challenges in the analysis of real-traffic data from naturalistic studies. 

• Understand how technologies and AI-tools may influence analysis and modelling of road-user behaviour in the future. 

• Compare the currently available tools for the virtual evaluation of active safety systems.