Reginald Roy Souleyrette
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Reginald Roy Souleyrette has a total of 37 co-authors. Here are the top co-authors by number of shared publications:
- Teng Wang
Kentucky Transportation Center, University of Kentucky, Lexington, KY 40506, USA
9 shared publications - Mohammad Zana Majidi
Department of Civil Engineering, University of Kentucky, Lexington, KY 40506, USA
4 shared publications - Antonio Pratelli
Department of Civil and Industrial Engineering, University of Pisa, Largo Lucio Lazzarino 1, 56122 Pisa, Italy
3 shared publications - Long Chen
Key Laboratory of Transport Industry of Big Data Application Technologies for Comprehensive Transport, Beijing Jiaotong University, Beijing, People’s Republic of China
3 shared publications - Robert Kluger
Department of Civil and Environmental Engineering, University of Louisville, Louisville, KY
3 shared publications - Lorenzo Brocchini
Department of Civil and Industrial Engineering, University of Pisa, Largo Lucio Lazzarino 1, 56122 Pisa, Italy
2 shared publications
Latest Publications
- 2 November 2026
- Published by Elsevier in Journal of Safety Research
- Vol. 99, 440-451
Quantifying systematic bias in police injury severity classifications: An ordinal modeling approach with linked Crash-Trauma records
- 2 November 2026
- Published by Elsevier in Journal of Safety Research
- Vol. 99, 440-451
Quantifying systematic bias in police injury severity classifications: An ordinal modeling approach with linked Crash-Trauma records
- Open Access
- 16 April 2026
- Published by MDPI AG in Urban Science
- Vol. 10 (4) , 215
Unconventional Roundabouts: Third-Generation Insights from the United States and Europe
This paper presents a comparative analysis of unconventional roundabouts, commonly referred to as third-generation roundabouts, based on case studies from the United States and Europe, specifically Croatia and Italy. These intersection designs deviate from traditional circular geometries to overcome limitations in safety, capacity, and spatial integration, especially in constrained or high-demand environments. The study focuses on three major typologies: raindrop, turbo, and two-geometry roundabouts (TGRs), examining their geometric characteristics, operational principles, and context-specific implementations. Based on real-world examples and qualitative assessments, each national section investigates design rationale and performance considerations in relation to local traffic dynamics. The paper contributes to the understanding of adaptive and context-sensitive intersection design, offering a conceptual framework for comparing unconventional roundabout typologies across different regulatory environments. The comparative analysis reveals that each typology responds to specific operational and spatial constraints rather than representing a universally optimal solution. In particular, raindrop and dog-bone roundabouts are most effective in interchange contexts, turbo roundabouts enhance safety and capacity in regulated multilane environments, while TGRs provide greater adaptability in constrained and irregular urban settings. These findings highlight the importance of context-dependent design strategies and support the need for flexible and context-sensitive evaluation frameworks.
- 16 April 2026
- Published by MDPI AG in Urban Science
- Vol. 10 (4) , 215
Open AccessUnconventional Roundabouts: Third-Generation Insights from the United States and Europe
This paper presents a comparative analysis of unconventional roundabouts, commonly referred to as third-generation roundabouts, based on case studies from the United States and Europe, specifically Croatia and Italy. These intersection designs deviate from traditional circular geometries to overcome limitations in safety, capacity, and spatial integration, especially in constrained or high-demand environments. The study focuses on three major typologies: raindrop, turbo, and two-geometry roundabouts (TGRs), examining their geometric characteristics, operational principles, and context-specific implementations. Based on real-world examples and qualitative assessments, each national section investigates design rationale and performance considerations in relation to local traffic dynamics. The paper contributes to the understanding of adaptive and context-sensitive intersection design, offering a conceptual framework for comparing unconventional roundabout typologies across different regulatory environments. The comparative analysis reveals that each typology responds to specific operational and spatial constraints rather than representing a universally optimal solution. In particular, raindrop and dog-bone roundabouts are most effective in interchange contexts, turbo roundabouts enhance safety and capacity in regulated multilane environments, while TGRs provide greater adaptability in constrained and irregular urban settings. These findings highlight the importance of context-dependent design strategies and support the need for flexible and context-sensitive evaluation frameworks.
- Open Access
- 4 December 2025
- Published by MDPI AG in Future Transportation
- Vol. 5 (4) , 190
From Prediction to Prevention: Identifying Actionable Crash Factors Through ML and Narrative-Based Sensitivity Testing
Crashes on roadways continue to represent a major global public health concern due to high rates of death and injury, underscoring the need for predictive tools that can identify high-risk conditions and guide prevention strategies. This study develops a framework that combines structured crash records and road information with unstructured police narratives to predict injury severity using machine learning and natural language processing (NLP). The dataset is used to train, validate, and test nine models, combining three algorithms (Random Forest, AdaBoost, and XGBoost) with two NLP methods (TF-IDF and Word2Vec). Model performance is evaluated using macro-average F1-scores to address severe class imbalance. Results show that XGBoost with TF-IDF achieves the best performance (macro-F1 = 0.644), demonstrating measurable improvements from incorporating narrative features compared to structured data alone. Beyond prediction, a simulation-based sensitivity analysis is conducted on the top 100 features, identifying 11 variables with the greatest impact on severity outcomes in Kentucky. Seatbelt non-use, occupant entrapment, and impaired driver control emerge as the most influential factors, with simulated improvements leading to notable reductions in fatalities and major injuries. The study introduces a “prediction-to-prevention” framework that links injury severity prediction with simulation-based sensitivity analysis. By integrating structured and narrative crash data, the framework identifies how changes in key behavioral and roadway factors can shift injury outcomes toward less severe levels. These findings highlight the dual contribution of this study: improving predictive accuracy through narrative integration and offering actionable insights to support evidence-based traffic safety interventions.
- 4 December 2025
- Published by MDPI AG in Future Transportation
- Vol. 5 (4) , 190
Open AccessFrom Prediction to Prevention: Identifying Actionable Crash Factors Through ML and Narrative-Based Sensitivity Testing
Crashes on roadways continue to represent a major global public health concern due to high rates of death and injury, underscoring the need for predictive tools that can identify high-risk conditions and guide prevention strategies. This study develops a framework that combines structured crash records and road information with unstructured police narratives to predict injury severity using machine learning and natural language processing (NLP). The dataset is used to train, validate, and test nine models, combining three algorithms (Random Forest, AdaBoost, and XGBoost) with two NLP methods (TF-IDF and Word2Vec). Model performance is evaluated using macro-average F1-scores to address severe class imbalance. Results show that XGBoost with TF-IDF achieves the best performance (macro-F1 = 0.644), demonstrating measurable improvements from incorporating narrative features compared to structured data alone. Beyond prediction, a simulation-based sensitivity analysis is conducted on the top 100 features, identifying 11 variables with the greatest impact on severity outcomes in Kentucky. Seatbelt non-use, occupant entrapment, and impaired driver control emerge as the most influential factors, with simulated improvements leading to notable reductions in fatalities and major injuries. The study introduces a “prediction-to-prevention” framework that links injury severity prediction with simulation-based sensitivity analysis. By integrating structured and narrative crash data, the framework identifies how changes in key behavioral and roadway factors can shift injury outcomes toward less severe levels. These findings highlight the dual contribution of this study: improving predictive accuracy through narrative integration and offering actionable insights to support evidence-based traffic safety interventions.
- Open Access
- 16 September 2025
- Published by MDPI AG in Preprints.org
From Prediction to Prevention: Identifying Actionable Crash Factors Through ML and Narrative-Based Sensitivity Testing
Traffic safety remains a critical area in transportation research due to its direct impact on public health. Reducing crash-related injuries and fatalities requires identifying key factors influencing injury severity. This study evaluates the predictive performance of three machine learning (ML) models; Random Forest (RF), XGBoost, and AdaBoost using a combination of structured crash data, highway characteristics, and unstructured crash narratives from police reports in Kentucky (2019–2023). RF was first applied for feature selection, identifying the top 100 predictors from over 350 variables. Two natural language processing (NLP) techniques; Term Frequency–Inverse Document Frequency (TF-IDF) and Word2Vec were then used to extract meaning from the crash narratives. These textual features were integrated with structured data to develop individual-level injury severity prediction models. XGBoost combined with TF-IDF yielded the highest predictive performance. The model was trained on 32,000 labeled records and then applied to a larger dataset of 67,000 crashes, many of which lacked injury severity data. The predicted results served as a baseline (Rb) for simulation-based sensitivity analysis. In this step, key input variables were systematically modified, and new predictions (R1, R2, etc.) were compared to assess how severity distributions shifted across four classes: fatal, major, minor, and possible injury. The analysis revealed that occupant entrapment, restraint use, not under proper control, ejection, speeding, and helmet use significantly influenced injury severity shifting. Changes in these factors often shifted outcomes from severe to less severe categories. These findings offer actionable insights for safety-focused policy and intervention strategies.
- 16 September 2025
- Published by MDPI AG in Preprints.org
Open AccessFrom Prediction to Prevention: Identifying Actionable Crash Factors Through ML and Narrative-Based Sensitivity Testing
Traffic safety remains a critical area in transportation research due to its direct impact on public health. Reducing crash-related injuries and fatalities requires identifying key factors influencing injury severity. This study evaluates the predictive performance of three machine learning (ML) models; Random Forest (RF), XGBoost, and AdaBoost using a combination of structured crash data, highway characteristics, and unstructured crash narratives from police reports in Kentucky (2019–2023). RF was first applied for feature selection, identifying the top 100 predictors from over 350 variables. Two natural language processing (NLP) techniques; Term Frequency–Inverse Document Frequency (TF-IDF) and Word2Vec were then used to extract meaning from the crash narratives. These textual features were integrated with structured data to develop individual-level injury severity prediction models. XGBoost combined with TF-IDF yielded the highest predictive performance. The model was trained on 32,000 labeled records and then applied to a larger dataset of 67,000 crashes, many of which lacked injury severity data. The predicted results served as a baseline (Rb) for simulation-based sensitivity analysis. In this step, key input variables were systematically modified, and new predictions (R1, R2, etc.) were compared to assess how severity distributions shifted across four classes: fatal, major, minor, and possible injury. The analysis revealed that occupant entrapment, restraint use, not under proper control, ejection, speeding, and helmet use significantly influenced injury severity shifting. Changes in these factors often shifted outcomes from severe to less severe categories. These findings offer actionable insights for safety-focused policy and intervention strategies.
- Open Access
Predicting person-level injury severity using crash narratives: A balanced approach with roadway classification and natural language process techniques
Predicting injuries and fatalities in traffic crashes plays a critical role in enhancing road safety, improving emergency response, and guiding public health interventions. This study investigates the added value of unstructured crash narratives (written by police officers at the scene) when combined with structured crash data to predict injury severity. Two widely used Natural Language Processing (NLP) techniques, Term Frequency-Inverse Document Frequency (TF-IDF) and Word2Vec, were employed to extract semantic meaning from the narratives, and their effectiveness was compared. To address the challenge of class imbalance, a K-Nearest Neighbors-based oversampling method was applied to the training data prior to modeling. The dataset consists of crash records from Kentucky spanning 2019 to 2023. To account for roadway heterogeneity, three road classification schemes were used: (1) eight detailed functional classes (e.g., Urban Two-Lane, Rural Interstate, Urban Multilane Divided), (2) four broader paired categories (e.g., Urban vs. Rural, Freeway vs. Non-Freeway), and (3) a unified dataset without classification. A total of 102 machine learning models were developed by combining structured features and narrative-based features using the two NLP techniques alongside three ensemble algorithms: XGBoost, Random Forest, and AdaBoost. Results demonstrate that models incorporating narrative data consistently outperform those relying solely on structured data. Among all combinations, TF-IDF coupled with XGBoost yielded the most accurate predictions in most subgroups. The findings highlight the power of integrating textual and structured crash information to enhance person-level injury prediction. This work offers a practical and adaptable framework for transportation safety professionals to improve crash severity modeling, guide policy decisions, and design more effective countermeasures.
Open AccessPredicting person-level injury severity using crash narratives: A balanced approach with roadway classification and natural language process techniques
Predicting injuries and fatalities in traffic crashes plays a critical role in enhancing road safety, improving emergency response, and guiding public health interventions. This study investigates the added value of unstructured crash narratives (written by police officers at the scene) when combined with structured crash data to predict injury severity. Two widely used Natural Language Processing (NLP) techniques, Term Frequency-Inverse Document Frequency (TF-IDF) and Word2Vec, were employed to extract semantic meaning from the narratives, and their effectiveness was compared. To address the challenge of class imbalance, a K-Nearest Neighbors-based oversampling method was applied to the training data prior to modeling. The dataset consists of crash records from Kentucky spanning 2019 to 2023. To account for roadway heterogeneity, three road classification schemes were used: (1) eight detailed functional classes (e.g., Urban Two-Lane, Rural Interstate, Urban Multilane Divided), (2) four broader paired categories (e.g., Urban vs. Rural, Freeway vs. Non-Freeway), and (3) a unified dataset without classification. A total of 102 machine learning models were developed by combining structured features and narrative-based features using the two NLP techniques alongside three ensemble algorithms: XGBoost, Random Forest, and AdaBoost. Results demonstrate that models incorporating narrative data consistently outperform those relying solely on structured data. Among all combinations, TF-IDF coupled with XGBoost yielded the most accurate predictions in most subgroups. The findings highlight the power of integrating textual and structured crash information to enhance person-level injury prediction. This work offers a practical and adaptable framework for transportation safety professionals to improve crash severity modeling, guide policy decisions, and design more effective countermeasures.
- Open Access
- 5 June 2025
- Published by Faculty of Transport and Traffic Sciences in Promet - Traffic&Transportation
- Vol. 37 (3) , 598-612
A Mixed Method Approach for Modelling Entry Capacity at Rotary Intersections
Rotary intersections, known as old traffic circles, require vehicles entering from branches to yield to circulating traffic. Upon entering, vehicles travel around a central island and exit toward their desired branch, generating merging and diverging conflicts at entry and exit points. Rotary capacity models are focused on the weaving manoeuvres within the circular roadway sections, associating capacity with the maximum traffic flow rate of each weaving segment. This paper introduces a novel approach combining modern roundabouts capacity models with the old rotary ones. In particular, the present study proposes a mixed approach based on an iterative process that combines the English TRRL model, which is suited for the old rotaries and based on short weaving sections capacity, with the features of the HCM-7th entry capacity model of the modern roundabouts, which is based on the circulating-traffic priority rule. Such an approach is rooted in the total capacity criteria and traffic conditions where all roundabout entrances reach congestion simultaneously. Compared to the past, this new approach makes entry performance estimation, such as average delay and queue length, bridging the gap between outdated and current methodologies in the field of rotary intersection design and assessment.
- 5 June 2025
- Published by Faculty of Transport and Traffic Sciences in Promet - Traffic&Transportation
- Vol. 37 (3) , 598-612
Open AccessA Mixed Method Approach for Modelling Entry Capacity at Rotary Intersections
Rotary intersections, known as old traffic circles, require vehicles entering from branches to yield to circulating traffic. Upon entering, vehicles travel around a central island and exit toward their desired branch, generating merging and diverging conflicts at entry and exit points. Rotary capacity models are focused on the weaving manoeuvres within the circular roadway sections, associating capacity with the maximum traffic flow rate of each weaving segment. This paper introduces a novel approach combining modern roundabouts capacity models with the old rotary ones. In particular, the present study proposes a mixed approach based on an iterative process that combines the English TRRL model, which is suited for the old rotaries and based on short weaving sections capacity, with the features of the HCM-7th entry capacity model of the modern roundabouts, which is based on the circulating-traffic priority rule. Such an approach is rooted in the total capacity criteria and traffic conditions where all roundabout entrances reach congestion simultaneously. Compared to the past, this new approach makes entry performance estimation, such as average delay and queue length, bridging the gap between outdated and current methodologies in the field of rotary intersection design and assessment.
- Open Access
- 4 March 2025
- Published by MDPI AG in Future Transportation
- Vol. 5 (1) , 30
Assessing Safety Performance of Complete Streets Projects
Complete Streets (CS) are defined as streets that accommodate all types of users, regardless of ability, safely and equitably allowing for the presence of pedestrians, bicyclists, transit users, and vehicle drivers to share the roadway. Several agencies have developed CS policies as a vital strategy to create more inclusive and accessible environments for all road users. CS are an efficient way to support the implementation of a multimodal transportation system, providing alternatives to car-oriented roadway designs. The Kentucky Transportation Cabinet recently developed the Complete Streets, Roads, and Highways Manual, aiming to implement a safe and equitable transportation system throughout the state. However, there is a need to evaluate the benefits of CS regarding their safety performance. This study aims to present crash data and summary statistics for CS projects that have been completed in Kentucky. The methodology involves a comparative analysis of safety data collected before and after the implementation of these projects. The results reveal that CS can be an effective approach to improve safety for all road users, including vulnerable and motor vehicle users. The findings also contribute to the existing knowledge on CS, offering insights into their impact on safety performance. Given that transportation agencies continue to prioritize sustainable and inclusive transportation solutions, the outcomes of this study will provide practical guidance for urban planners, policymakers, and transportation engineers seeking evidence-based solutions for creating safer roads.
- 4 March 2025
- Published by MDPI AG in Future Transportation
- Vol. 5 (1) , 30
Open AccessAssessing Safety Performance of Complete Streets Projects
Complete Streets (CS) are defined as streets that accommodate all types of users, regardless of ability, safely and equitably allowing for the presence of pedestrians, bicyclists, transit users, and vehicle drivers to share the roadway. Several agencies have developed CS policies as a vital strategy to create more inclusive and accessible environments for all road users. CS are an efficient way to support the implementation of a multimodal transportation system, providing alternatives to car-oriented roadway designs. The Kentucky Transportation Cabinet recently developed the Complete Streets, Roads, and Highways Manual, aiming to implement a safe and equitable transportation system throughout the state. However, there is a need to evaluate the benefits of CS regarding their safety performance. This study aims to present crash data and summary statistics for CS projects that have been completed in Kentucky. The methodology involves a comparative analysis of safety data collected before and after the implementation of these projects. The results reveal that CS can be an effective approach to improve safety for all road users, including vulnerable and motor vehicle users. The findings also contribute to the existing knowledge on CS, offering insights into their impact on safety performance. Given that transportation agencies continue to prioritize sustainable and inclusive transportation solutions, the outcomes of this study will provide practical guidance for urban planners, policymakers, and transportation engineers seeking evidence-based solutions for creating safer roads.
- 9 January 2024
- Published by Elsevier in Accident Analysis & Prevention
- Vol. 197, 107461
A systematic review and meta-analysis of data linkage between motor vehicle crash and hospital-based datasets
- 9 January 2024
- Published by Elsevier in Accident Analysis & Prevention
- Vol. 197, 107461
A systematic review and meta-analysis of data linkage between motor vehicle crash and hospital-based datasets
- 8 July 2023
- Published by Elsevier in Computers & Operations Research
- Vol. 159
An improved multi-objective framework for the Rich arc routing problem
- 8 July 2023
- Published by Elsevier in Computers & Operations Research
- Vol. 159
An improved multi-objective framework for the Rich arc routing problem