Project Description

While safety is a paramount objective for the whole transport system, road safety deserves particular attention, given the disproportionate occurrence of accidents and casualties. The European Commission sets ambitious targets: eliminating road crashes caused by human error by 2050 and ensuring that the EU remains a world leader in transport safety. The “Vision Zero” strategy is heavily reliant on vehicle automation, which is expected to take the human driver “out of the loop” at its most advanced stage of development. However, as full automation is not immediately achievable, the EU Road Safety Policy Framework 2021-2030 established a range of key performance indicators (KPIs) for road safety on the way to “Vision Zero”. Being the human error involved in some way in 95% of road crashes, either as the main cause or as a contributing factor, KPIs to assess progress on safe driving behaviours, such as speed and driving impairment, are naturally defined. After years of strong investments in infrastructures, law enforcement and driver training/education, in which the EU observed a steady decrease of road deaths and serious injuries, the efforts to further reduce those figures are now shifting to emerging monitoring and warning technologies. This allows to act proactively in the prevention of crashes by providing real-time feedback to the driver about his/her driving performance and physiological state, or about changing road conditions. In fact, during the transition to full automation, driver assistance and driver/road monitoring systems will continue to be developed to become more reliable and compliant with drivers’ expectations.

uRisk will contribute to this endeavour by conducting an exploratory study on human errors and risky behaviours at the wheel. The project aims to deliver a generalised index to characterise unsafe driving that will combine multiple dynamic driving and driver’s physiological parameters. Variations in these parameters are associated with the level of risk at the wheel, and consequently, with the occurrence of traffic incidents. The index will incorporate and hierarchize these measures in a standardised manner. Any traffic incident or accident is preceded by variations in dynamic driving parameters (e.g., trajectory, speed, acceleration, and following distance), which are not adequate to the local conditions of the road, traffic, or surrounding environment. These are accompanied by reactions in the driver’s physiological signals (e.g., heart rate, heart rate variability, brain activity, and eye movement). Sudden variations in dynamic driving and biometric parameters are many times the physical consequence of human errors, which may be produced by instant or permanent risky situations. The instant risk factors include the driver’s response to the road geometric and environmental features, traffic conditions, distractions, and unexpected events. The permanent risk factors are mainly related to the driver’s characteristics or state (e.g., gender, age, physical conditions, and psychological conditions), and also to travel motivation. The new index can be embedded into new in-vehicle driver monitoring/warning systems, improving human-systems integration and reducing the probability of “false alerts” that affect users’ trust and compliance with the systems’ recommendations.

uRisk will explore the sequential and combined cause-effect relations between risk factors, physiological signals and variations of driving parameters in traffic incidents by conducting driving experiments in a simulated environment. The research team will test a set of risk factors as comprehensive as possible with simultaneous monitoring of dynamic driving parameters and physiological signals in a driving simulator. Nonintrusive equipment, such as eye-tracking and heart-rate monitoring devices, will be used to assess the variations in the driver’s physiological signals. The data obtained through the experiments will be analysed and processed through advanced statistical and machine learning techniques to predict the evolution of drivers’ behaviour and/or state based on the generalised unsafe driving index.

In the end, uRisk will contribute to the state of the art by integrating information about the driver with behaviour factors and increased awareness on risky behaviours and recommendations on new solutions to improve road safety by reducing the risk and impact of human errors, aiming to support the authorities in policy-making and the industry in developing enhanced ICT solutions, automation, and Advanced Driver Assistance Systems (ADAS).

Research Team

· Sérgio Pedro Duarte (PI)

· Carlos Rodrigues

· José Pedro Tavares

· Sara Ferreira

Financial Support

· 49.938,81 EUR

Stage of Progress

· Selected for funding (expected to start in January 2025)