University of Aveiro

The University of Aveiro (UA) has established expertise in transport, mobility and road-safety research, combining engineering, data science, mathematical modelling and policy-oriented approaches to address complex transport challenges. Through its research centres and multidisciplinary teams, UA develops research that connects scientific knowledge with real-world mobility systems and supports evidence-based decision-making by public authorities and other stakeholders. The University has an extensive portfolio of national and international research and technology-transfer projects, reflecting its capacity to collaborate with public institutions, companies and international partners.

Within the field of road safety, UA has a longstanding collaboration with Portuguese public authorities, notably the National Road Safety Authority (ANSR) and the Mobility and Transport Institute (IMT). This collaboration has enabled researchers to work with real-world road-crash and mobility data and to translate scientific evidence into knowledge that can support road-safety assessment, planning and policy development. In particular, collaboration with ANSR has provided access to detailed crash databases used in research on crash occurrence, injury severity and vulnerable road users.

Research conducted at UA by Eloísa Macedo, in collaboration with researchers in transport and road-safety engineering, has contributed to the development of quantitative approaches for understanding and predicting road-crash outcomes. This research includes the identification of factors influencing crash and injury severity, the development of statistical and machine-learning models for predicting crash severity, and the analysis of spatial and temporal patterns of road crashes. For example, research involving Macedo developed multinomial logistic regression approaches to identify factors associated with the occurrence and severity of crashes involving vulnerable road users (VRUs), while spatial analysis was used to identify high-risk locations and patterns. The results demonstrated how spatial and temporal variables can be incorporated into predictive models to support the formulation of road-safety policies.

A particular research focus has been placed on vulnerable road users, including pedestrians and cyclists. Studies using crash data from Aveiro investigated the characteristics and risk factors associated with crashes involving VRUs, including environmental conditions, crash location, pedestrian crossings and characteristics of the road users. More recent work has extended this research towards emerging forms of micromobility, including e-scooters, combining national crash data with real-world experiments in Aveiro to evaluate safety, comfort and operational performance.

The research portfolio also encompasses crash injury-severity modelling and risk prediction, including approaches capable of addressing highly unbalanced crash datasets and rare but severe outcomes. Machine-learning and statistical techniques have been investigated to identify the variables that contribute to increased injury severity and to improve the prediction of severe outcomes. Such approaches provide an important bridge between retrospective crash analysis and proactive road-safety management, allowing authorities to identify risk factors, prioritise interventions and assess potential countermeasures.

The University of Aveiro therefore provides a strong research-to-policy environment for road safety, in which empirical crash data, statistical modelling, artificial intelligence, spatial analysis and transport engineering are integrated to generate actionable evidence. The combination of long-term collaboration with Portuguese road-safety and transport authorities, access to real-world data, multidisciplinary research expertise and experience in translating analytical results into policy-relevant knowledge positions UA to contribute effectively to data-driven road-safety strategies, risk assessment, vulnerable-road-user protection and the development of evidence-based interventions.