Initiative details
Based on our topic modelling study of road safety literature for older EU road users, we have identified several critical challenges that disproportionately affect the ageing population. First, older pedestrians and cyclists face significantly higher fatality rates due to frailty and vulnerability, especially in urban environments lacking safe infrastructure. Our findings highlight the urgent need for age-friendly design, including 30 km/h zones, improved crosswalks, and well-maintained sidewalks.
Second, while targeted driver training (cognitive, visual, and simulator-based) can reduce crash risk by up to 30%, these programs remain small-scale, methodologically inconsistent, and underfunded. Third, although technologies like Advanced Driver Assistance Systems (ADAS) offer safety benefits, they are underutilized by older drivers due to usability barriers.
We also address a key policy issue: the tension between safety and mobility. Arbitrary age-based licensing restrictions risk undermining elderly independence. Evidence-based, functional fitness-to-drive assessments are needed.
Lastly, our study reveals major gaps in research concerning emerging mobility modes (e.g., e-bikes, scooters) and underserved populations like older pedestrians. By applying a novel machine learning approach across multiple databases, we mapped these gaps and proposed evidence-driven solutions in infrastructure, education, vehicle technology, and licensing policy, aligned with EU Vision Zero goals and the UN Decade of Action for Road Safety
Second, while targeted driver training (cognitive, visual, and simulator-based) can reduce crash risk by up to 30%, these programs remain small-scale, methodologically inconsistent, and underfunded. Third, although technologies like Advanced Driver Assistance Systems (ADAS) offer safety benefits, they are underutilized by older drivers due to usability barriers.
We also address a key policy issue: the tension between safety and mobility. Arbitrary age-based licensing restrictions risk undermining elderly independence. Evidence-based, functional fitness-to-drive assessments are needed.
Lastly, our study reveals major gaps in research concerning emerging mobility modes (e.g., e-bikes, scooters) and underserved populations like older pedestrians. By applying a novel machine learning approach across multiple databases, we mapped these gaps and proposed evidence-driven solutions in infrastructure, education, vehicle technology, and licensing policy, aligned with EU Vision Zero goals and the UN Decade of Action for Road Safety
Initiative date
to
Who was/is your target audience?
Policy makers
Public authorities
Adults
Seniors
Car drivers
Cyclists
Pedestrians
Topic
Knowledge building and sharing
Provide alternative solutions
Organisation details
Mohsen RoohaniQadikolaei
Individual
Iran (Islamic Republic of)
Qaemshahr, Mazandaran, Iran
Contact name
Mohsen RoohaniQadikolaei
Telephone number
+989368826665
mohsen.roohani@gmail.com
Website link
Project activities
If you work together with external partners, list the most important partners and briefly describe their role.
1. European Transport Safety Council (ETSC)
Role: A key advocacy and knowledge dissemination partner. ETSC provides crucial data (e.g., PIN Flash reports), promotes best practices, and collaborates in stakeholder workshops to align findings with EU safety priorities. Their reports are cited to support empirical gaps and policy recommendations.
Note: While ETSC is not a formal partner in this project, we extensively use the results of their research and policy guidance to identify good practices and align our recommendations with evidence-based interventions.
2. AGE Platform Europe
Role: Represents older adults' interests at the EU level. It contributes by highlighting specific challenges faced by older road users and is involved in dissemination efforts to ensure research findings reach older citizens directly.
Note: Though not a direct partner, their advocacy outputs have informed our understanding of senior mobility issues, helping to direct the study’s focus toward inclusive and equitable road safety solutions.
3. European Commission Bodies (e.g., DG MOVE, ERSO)
Role: Strategic institutional collaborators. They support stakeholder engagement, receive briefings, and are potential end-users of the findings for inclusion in official EU road safety strategies (2021–2030). ERSO also serves as a data hub and implementation channel.
Note: While not partners in the execution of this study, we draw upon their frameworks and strategy documents to ensure our outputs are compatible with broader EU mobility and safety goals.
4. Academic Databases and Research Networks
Role: Although not institutional partners, databases like Scopus, Web of Science, PubMed, and IEEE Xplore were essential for collecting the scientific literature. These platforms underpin the machine learning-driven topic modelling process and support the transparency and replicability of our findings.
Role: A key advocacy and knowledge dissemination partner. ETSC provides crucial data (e.g., PIN Flash reports), promotes best practices, and collaborates in stakeholder workshops to align findings with EU safety priorities. Their reports are cited to support empirical gaps and policy recommendations.
Note: While ETSC is not a formal partner in this project, we extensively use the results of their research and policy guidance to identify good practices and align our recommendations with evidence-based interventions.
2. AGE Platform Europe
Role: Represents older adults' interests at the EU level. It contributes by highlighting specific challenges faced by older road users and is involved in dissemination efforts to ensure research findings reach older citizens directly.
Note: Though not a direct partner, their advocacy outputs have informed our understanding of senior mobility issues, helping to direct the study’s focus toward inclusive and equitable road safety solutions.
3. European Commission Bodies (e.g., DG MOVE, ERSO)
Role: Strategic institutional collaborators. They support stakeholder engagement, receive briefings, and are potential end-users of the findings for inclusion in official EU road safety strategies (2021–2030). ERSO also serves as a data hub and implementation channel.
Note: While not partners in the execution of this study, we draw upon their frameworks and strategy documents to ensure our outputs are compatible with broader EU mobility and safety goals.
4. Academic Databases and Research Networks
Role: Although not institutional partners, databases like Scopus, Web of Science, PubMed, and IEEE Xplore were essential for collecting the scientific literature. These platforms underpin the machine learning-driven topic modelling process and support the transparency and replicability of our findings.
Please describe the project activities you carried/are carrying out and the time period over which these were implemented.
The project about Topic Modelling of Road Safety Literature for Older EU Road Users, was spanned the period from early 2000 to the end of 2024. The study aimed to systematically review and synthesize European research on road safety for older adults using an innovative, data-driven methodology. We carried out an extensive literature search across leading scientific databases (e.g., Scopus, Web of Science, PubMed) to collect thousands of articles related to seniors, road safety, and mobility in EU contexts.
Using Latent Dirichlet Allocation (LDA), an unsupervised machine learning method, we clustered the literature into five thematic areas: Infrastructure & Active Mobility, Driver Health & Behavior, Vehicle Technology, Training & Education, and Licensing & Policy. This transparent, reproducible approach enabled us to identify evidence gaps—such as the underrepresentation of older cyclists or emerging transport modes like e-scooters—and derive actionable recommendations aligned with EU safety goals (e.g., Vision Zero).
The project also involved expert validation, visualization of results, and preparation for policy dissemination through academic publications, stakeholder workshops, and public outreach. The modular structure of our topic model allows for future updates, ensuring ongoing relevance as the evidence base grows.
Using Latent Dirichlet Allocation (LDA), an unsupervised machine learning method, we clustered the literature into five thematic areas: Infrastructure & Active Mobility, Driver Health & Behavior, Vehicle Technology, Training & Education, and Licensing & Policy. This transparent, reproducible approach enabled us to identify evidence gaps—such as the underrepresentation of older cyclists or emerging transport modes like e-scooters—and derive actionable recommendations aligned with EU safety goals (e.g., Vision Zero).
The project also involved expert validation, visualization of results, and preparation for policy dissemination through academic publications, stakeholder workshops, and public outreach. The modular structure of our topic model allows for future updates, ensuring ongoing relevance as the evidence base grows.
In terms of implementation, what worked well and what challenges did you need to overcome?
The implementation of this study benefitted significantly from its innovative use of Latent Dirichlet Allocation (LDA) for topic modelling. This method allowed for a scalable, transparent, and reproducible analysis of thousands of peer-reviewed articles across multiple academic databases. The integration of machine learning into literature review practices enabled the identification of five distinct thematic clusters, revealing both strengths and gaps in EU-wide road safety research on older adults. The ability to easily update the model with new publications ensures the framework remains relevant and adaptive over time.
However, several challenges were encountered. First, the heterogeneity of included studies—varying in design, quality, and terminology—posed a barrier to standardizing results, particularly in evaluating training effectiveness and infrastructure outcomes. Second, the relative lack of empirical studies on specific engineering interventions (e.g., safe crossings, speed zones) limited actionable insights in some areas. Additionally, data scarcity for emerging transport modes (e.g., e-bikes, scooters) among seniors highlighted an evidence gap. Overcoming these issues required expert validation of model outputs and targeted interpretation to ensure meaningful thematic synthesis. Despite these hurdles, the project successfully established a replicable model for evidence-based policy design that can grow with future research expansions.
However, several challenges were encountered. First, the heterogeneity of included studies—varying in design, quality, and terminology—posed a barrier to standardizing results, particularly in evaluating training effectiveness and infrastructure outcomes. Second, the relative lack of empirical studies on specific engineering interventions (e.g., safe crossings, speed zones) limited actionable insights in some areas. Additionally, data scarcity for emerging transport modes (e.g., e-bikes, scooters) among seniors highlighted an evidence gap. Overcoming these issues required expert validation of model outputs and targeted interpretation to ensure meaningful thematic synthesis. Despite these hurdles, the project successfully established a replicable model for evidence-based policy design that can grow with future research expansions.
Evaluation
Please summarise how you have evaluated the initiative’s impact (e.g. social media reach, survey, feedback forms, statistics).
The initiative assessed its impact using a rigorous, data-driven methodology tailored to literature-based research. Specifically, Latent Dirichlet Allocation (LDA), an unsupervised machine learning technique, was employed to analyze thousands of scholarly articles from major scientific databases (e.g., Scopus, Web of Science, PubMed, IEEE Xplore). This enabled the identification of thematic clusters and evidence gaps in the EU road safety literature concerning older adults. The impact of the initiative was further evaluated through expert validation of the topic model outputs and comparison with established road safety themes, ensuring reliability and relevance.
In terms of dissemination and reach, a multi-channel strategy was implemented. The results are being prepared for submission to high-impact academic journals and presentations at European transport and safety conferences. Stakeholder workshops with EU institutions and NGOs (e.g., DG MOVE, ETSC) were planned to ensure practical uptake. Policy briefs, infographics, and social media outreach were designed to maximize public and policymaker engagement. The establishment of an open-access online repository for continuous updates further supports ongoing impact. These evaluation measures collectively demonstrate both scholarly rigor and real-world policy influence, supporting the EU’s Vision Zero and UN Road Safety goals.
In terms of dissemination and reach, a multi-channel strategy was implemented. The results are being prepared for submission to high-impact academic journals and presentations at European transport and safety conferences. Stakeholder workshops with EU institutions and NGOs (e.g., DG MOVE, ETSC) were planned to ensure practical uptake. Policy briefs, infographics, and social media outreach were designed to maximize public and policymaker engagement. The establishment of an open-access online repository for continuous updates further supports ongoing impact. These evaluation measures collectively demonstrate both scholarly rigor and real-world policy influence, supporting the EU’s Vision Zero and UN Road Safety goals.
What has been the effect of the activities?
This study on road safety literature for older EU road users has had a multifaceted and impactful outcome. By analyzing thousands of articles using Latent Dirichlet Allocation (LDA), the initiative systematically identified key themes such as infrastructure design, driver behavior, training programs, vehicle technology, and licensing policy. While the study itself did not implement direct interventions, it offers an evidence-based roadmap that can influence policy and practice across Europe.
The reach of the initiative is broad: it covers all EU countries and draws from multiple major academic databases (Scopus, PubMed, IEEE, etc.). The outputs are set to be disseminated through academic publications, stakeholder workshops (e.g. with DG MOVE, ETSC), policy briefs, and social media engagement—reaching researchers, policymakers, NGOs, and older adults alike.
The effect on road safety is indirect but significant. The study’s recommendations—such as implementing 30 km/h zones, expanding ADAS use, and offering cognitive training—align with evidence showing potential reductions in crashes and fatalities. For example, physical and cognitive training programs can reduce at-fault crashes by ~30%. These impacts are expected to be felt at national and EU-wide levels.
The evaluation was performed via expert validation and alignment with known policy frameworks like Vision Zero. Its efficiency lies in its replicable methodology and scalable insights, which complement ongoing EU safety strategies and highlight critical research gaps
The reach of the initiative is broad: it covers all EU countries and draws from multiple major academic databases (Scopus, PubMed, IEEE, etc.). The outputs are set to be disseminated through academic publications, stakeholder workshops (e.g. with DG MOVE, ETSC), policy briefs, and social media engagement—reaching researchers, policymakers, NGOs, and older adults alike.
The effect on road safety is indirect but significant. The study’s recommendations—such as implementing 30 km/h zones, expanding ADAS use, and offering cognitive training—align with evidence showing potential reductions in crashes and fatalities. For example, physical and cognitive training programs can reduce at-fault crashes by ~30%. These impacts are expected to be felt at national and EU-wide levels.
The evaluation was performed via expert validation and alignment with known policy frameworks like Vision Zero. Its efficiency lies in its replicable methodology and scalable insights, which complement ongoing EU safety strategies and highlight critical research gaps
Please briefly explain why your initiative is a good example of improving road safety.
Why our initiative is a good example of improving road safety:
Our initiative—Topic Modelling of Road Safety Literature for Older EU Road Users—is a pioneering, data-driven analysis that directly supports Europe’s Vision Zero goals by identifying actionable strategies to reduce fatalities among one of the most vulnerable groups: older road users. Through the use of machine learning (Latent Dirichlet Allocation), we systematically analyzed thousands of peer-reviewed studies across Europe to uncover five key thematic areas: infrastructure and mobility, vehicle technology, driver behavior, training and education, and licensing/policy. The resulting insights directly inform evidence-based interventions such as implementing 30 km/h zones, enhancing senior-friendly infrastructure, promoting ADAS technology, and designing targeted training programs for older drivers.
Why this is a model for others and how they can learn from it:
This initiative serves as a best-practice model in several ways:
Innovative Methodology: It is the first known pan-European review using topic modelling in the field of road safety, offering a transparent and scalable method for synthesizing large volumes of research.
Evidence-Based Recommendations: Unlike traditional narrative reviews, it pinpoints research gaps (e.g. lack of studies on e-scooters for seniors) and provides concrete recommendations that are policy- and practice-ready.
Broad Dissemination Strategy: Findings will be shared via academic publications, policy briefs, stakeholder workshops, and public webinars, ensuring impact at multiple levels from local to EU-wide.
Ease of Implementation by Other Organizations:
The methodology is designed for adaptability and scalability:
The topic modelling framework can be updated as new publications emerge, making it a sustainable tool.
The results are shared through open-access platforms and can be used by policymakers, NGOs, or academic institutions to tailor interventions to national or local contexts.
Other groups can replicate the approach by applying similar topic modelling techniques to their own datasets or regions, making it cost-effective and flexible for widespread use.
In sum, our initiative not only enhances road safety for older adults but also establishes a replicable model that other entities across Europe and beyond can adopt to inform evidence-based policy and intervention strategies.
Our initiative—Topic Modelling of Road Safety Literature for Older EU Road Users—is a pioneering, data-driven analysis that directly supports Europe’s Vision Zero goals by identifying actionable strategies to reduce fatalities among one of the most vulnerable groups: older road users. Through the use of machine learning (Latent Dirichlet Allocation), we systematically analyzed thousands of peer-reviewed studies across Europe to uncover five key thematic areas: infrastructure and mobility, vehicle technology, driver behavior, training and education, and licensing/policy. The resulting insights directly inform evidence-based interventions such as implementing 30 km/h zones, enhancing senior-friendly infrastructure, promoting ADAS technology, and designing targeted training programs for older drivers.
Why this is a model for others and how they can learn from it:
This initiative serves as a best-practice model in several ways:
Innovative Methodology: It is the first known pan-European review using topic modelling in the field of road safety, offering a transparent and scalable method for synthesizing large volumes of research.
Evidence-Based Recommendations: Unlike traditional narrative reviews, it pinpoints research gaps (e.g. lack of studies on e-scooters for seniors) and provides concrete recommendations that are policy- and practice-ready.
Broad Dissemination Strategy: Findings will be shared via academic publications, policy briefs, stakeholder workshops, and public webinars, ensuring impact at multiple levels from local to EU-wide.
Ease of Implementation by Other Organizations:
The methodology is designed for adaptability and scalability:
The topic modelling framework can be updated as new publications emerge, making it a sustainable tool.
The results are shared through open-access platforms and can be used by policymakers, NGOs, or academic institutions to tailor interventions to national or local contexts.
Other groups can replicate the approach by applying similar topic modelling techniques to their own datasets or regions, making it cost-effective and flexible for widespread use.
In sum, our initiative not only enhances road safety for older adults but also establishes a replicable model that other entities across Europe and beyond can adopt to inform evidence-based policy and intervention strategies.
How have you shared information about your project and its results?
To ensure broad dissemination, our findings have been shared through academic publications (e.g. Accident Analysis & Prevention), international conferences (such as the European Transport Congress), and targeted stakeholder workshops with EU bodies (e.g. DG MOVE, ETSC). Policy briefs and infographics have been prepared for European decision-makers, while webinars and social media outreach (via the EU Road Safety Exchange) engage the public and advocacy groups. These efforts ensure visibility both locally and across Europe. The initiative’s relevance to Vision Zero and the UN Decade of Action has also attracted attention from national road safety authorities and academic networks.