Road crashes remain among the gravest threats to public safety, and preventing them is a defining task of transportation systems worldwide. Much of that harm concentrates at hotspots, yet a hotspot is less a place than an episode; it emerges quietly at an intersection or along an arterial, intensifies for weeks, then subsides, only to reappear elsewhere. Enforcement guided by maps of past crashes inevitably trails this cycle, patrolling yesterday's hotspots while tomorrow's form unwatched. Breaking that lag requires three capabilities at once: detecting hotspots as they are born, forecasting where they will sit next week, and following each one through its life. We introduce HERALD (Hotspot Emergence, Risk Anticipation, and Life-cycle Dynamics), a unified deep learning framework that provides all three from a single statewide model. HERALD distills each county's recent crash history into weekly risk maps and forecasts the next with a CNN--Transformer, whose mixture-of-experts lets one model serve dense urban cores and sparse rural corridors alike. Each forecast is anchored in the county's long-run crash geography, sharpened by the self-exciting effect of recent crashes, and paired with explicit warnings of where new hotspots are about to appear. Followed over time, every hotspot acquires a legible life story, from birth through growth and stability to decline and death. Across six heterogeneous Wisconsin counties, HERALD forecasts more accurately than five identically trained baselines, locates hotspots most precisely, and flags emerging risks before they take hold. A single adjustable setting trades accuracy for extra sensitivity where deployment demands it. The result shifts hotspot management from mapping the past to anticipating the future.

The present study uses Crash Sequence Analysis to identify automated vehicles (AV) crash patterns and evaluate the temporal (2014 to 2023) and manufacturer-specific (Cruise or Waymo) trends in these patterns. This method builds upon the evaluation of crash scenarios by considering the sequential nature of crashes, incorporating crash progression and contributing factors. The results highlight the current challenge faced by manufacturers in designing systems that are safe yet perform in a manner expected by human drivers. The proportional reduction of crash patterns involving rear-end collisions during left or right turns suggests that actions were taken by manufacturers to address the aforementioned challenge. However, other crash patterns have shown a proportional increase in recent years, such as collisions with objects and on narrow roads, with specific variations by manufacturers. These findings provide key insights for shaping future AV development, guiding public sector decisions, and building public trust in automation.

Driving behavior and interactions with bicyclists on rural roads have not been quantified and modeled extensively. Naturalistic bicycling data for 1,991 passing events were collected on a rural two-lane roadway (55 mph, 88 kph speed limit) to quantify how opposing traffic and vehicle platooning influence passing lateral distance, speed, and aerodynamic forces. Results indicate that opposing traffic significantly reduces passing lateral distance by an average of 2.0 ft (61 cm) and decreases speed by an average of 2.3 mph (3.7 kph). Platooning leads to progressively reduced passing distance and speed among following vehicles. The reductions reflect limited available space and increased risk for bicyclists when opposing vehicles are present. The estimated aerodynamic lateral forces created by passenger vehicles were well below tolerable safety limits for bicyclists. To surpass tolerable limits, passenger vehicles would have to pass at a lateral distance of 0.9 ft (27 cm) at a speed of 55 mph (88 kph). Lateral distance and speed were found to be independent at a disaggregate level. Leading vehicles’ lateral distance followed a Log-normal distribution and speed followed a Weibull distribution. Theoretical joint probability density functions were developed for leading and following vehicles with and without opposing traffic. Pairwise differences among lead and follower vehicles were similar and resembled a Normal distribution. The developed joint probability density functions can be used for calibration and validation of driving simulators, or development of autonomous and artificial intelligence driving models. Results contribute to developing safer design guidance and risk mitigating strategies for bicyclists.

Understanding active transportation is critical for transportation planning, infrastructure development, and safety improvements. Unlike motor vehicles, which have widespread automated counting stations, cycling and walking automated counting has limited coverage. Given the limited data and unique characteristics of active transportation, it is crucial to evaluate the accuracy of counting technologies and account for temporal variations, weather effects, and transferability when estimating volumes. Data from four sites in Wisconsin were analyzed with 5 years of hourly sensor, weather, and Strava data, along with 268 h of manually processed ground truth video data. Ground truth hourly count trends showed that pedal cycles were the main users in the shared paths (78%-87%). There were peak and directional hourly trends by week or weekend days, higher volumes and a shift in the type of user were observed on weekends. Automatic sensor count data accuracy from inductive loop and infrared sensors was evaluated and compared with ground truth data. Inductive loop counting technology showed high levels of pedal cycle count accuracy (91%-92%). Infrared sensors counted passersby with a reduced degree of accuracy (54%-67%). Negative binomial regression modeling was implemented to account for overdispersion in the count data. Key predictors included time of day, day of the week, month, temperature, precipitation, and Strava counts. Site-specific models were developed, transferability across sites was assessed, and models were generalized with data from sites that shared similar characteristics applicable to high-volume, urban commuting and recreational paths. Models were not transferable to isolated sites with low volume and unreliable sensor count data.

Traffic demand varies significantly in urban areas, impacting intersection performance. Current control strategies assume fixed lane assignment with signal optimization that focuses on the traffic movements at each approach. This leads to inefficient use of temporal and spatial resources. To improve the intersection performance in such cases, Dynamic Lane Assignment (DLA), a component of Intelligent Transportation Systems (ITS), is employed to improve the intersection efficiency. As we move into an era where Connected and Automated Vehicle (CAV) technology is increasingly recognized for enhancing traffic safety and efficiency, it is important to consider that CAVs and human-driven vehicles (HDVs) will co-exist in the system for a substantial period. This necessitates managing CAVs alongside HDVs, which could benefit from dedicated lane assignments for CAVs together with signal optimization to enhance overall system performance and efficiency. However, research on CAV-based DLA combined with signal optimization in mixed traffic environments remains limited. This research focuses on understanding the state-of-the-art of managing intersections with DLA in mixed traffic environments. Information was gathered from previous research papers and public documents. For the synthesis, studies were categorized into the following topics: (1) application of DLA combined with signal optimization for CAVs; (2) application of DLA with CAVs for freeway management; and (3) other lane assignment strategies such as Dynamic Lane Grouping (DLG). The synthesis focuses on the assumptions, strategies, and policies related to lane assignment, as well as the methodologies employed in these studies. Finally, the paper provides a comprehensive overview of DLA strategies for intersection management, identifies research gaps, and proposes future directions for mixed traffic environments.

The Safe System Approach (SSA) aims to eliminate fatal and serious injury roadway crashes through a holistic view of the road system, moving away from traditional safety analysis based exclusively on historical crash data. One reason for this is the classification of crashes into broad categories (e.g., head-on, sideswipe), which does not capture crash progression or contributing factors. In this context, this paper applies crash sequence analysis to historical crash data and uses the findings to proactively identify safety issues in similar contexts, in alignment with the SSA framework. The method uses sequence-of-events information from crash data to generate clusters of crashes with similar underlying characteristics. Data from fatal and serious injury crashes from urban intersections in the state of Ohio between 2018 and 2022 were used in the analysis. The results show 12 clusters with unique characteristics that consider the sequence of events of each crash. Although derived from crash data, the clusters offer an in-depth understanding of the factors associated with each one and help identify cluster-specific countermeasures related to various SSA elements. State and local jurisdictions can use the presented methodology in transportation safety programs, by focusing on the clusters that represent local challenges or on countermeasures related to the issues of multiple clusters. Finally, the method can also be associated with site-specific analysis, providing a comprehensive toolkit for practitioners.

The goal of this research was to quantify the mobility and safety impacts of different combinations of lane width and shy distance to the barrier for a given paved width in work zones. The research team developed a device to measure lateral distance and derive speed, vehicle length/type, and headway information under day and night conditions. Data were collected at 17 locations in Illinois, Michigan, and Wisconsin. Lateral distance data of over a quarter million vehicles were used for the safety analysis. Extreme value theory modeling was conducted to estimate the probabilities of right-hand edge line encroachment and right-hand barrier contact. Wider lanes were found to have decreased probabilities of edge line encroachment and barrier contact, while wider shy distances were associated with increased probability of edge line encroachment and decreased probability of barrier contact. The speeds of over 125,000 free flow vehicles were used to quantify the mobility impact. Linear regression was implemented to develop models for estimating free flow speeds in work zones. Work zone free flow speed increased with an increase in speed limit, lane width, and left-/right-hand shy distance to the barrier. A case study of a 55-mph posted work zone with two open lanes and barrier on both sides with 26-ft available paved width is presented. Results of the case study indicate that 11-ft lanes with 2-ft shy distance have a slightly lower probability of right-hand barrier contact (for vehicles in the right-hand lane) than 12-ft lanes with 1-ft shy distance, while having a greater free flow speed. This research has demonstrated how lateral distance can be collected and modeled along with speed data to assess safety and mobility impacts in work zones.

Transportation Cyber-Physical Systems (T-CPS) enhance safety and mobility by integrating cyber and physical transportation systems. A key component of T-CPS is the Digital Twin (DT), a virtual representation that enables simulation, analysis, and optimization through real-time data exchange and communication. Although existing studies have explored DTs for vehicles, communications, pedestrians, and traffic, real-world validations and implementations of DTs that encompass infrastructure, vehicles, signals, and communications remain limited due to several challenges. These include accessing real-world connected infrastructure, integrating heterogeneous, multi-sourced data, ensuring real-time data processing, and synchronizing the digital and physical systems. To address these challenges, this study develops a traffic DT based on a real-world connected vehicle corridor. Leveraging the Cellular Vehicle-to-Everything (C-V2X) infrastructure in the corridor, along with communication, computing, and simulation technologies, the DT accurately replicates physical vehicle behaviors, signal timing, communications, and traffic patterns within the virtual environment. Building upon the previous data pipeline, the digital system ensures robust synchronization with the physical environment. Moreover, the DT’s scalable and redundant architecture enhances data integrity, making it capable of supporting future large-scale C-V2X deployments. Lastly, the DT’s ability to provide feedback to the physical system is demonstrated through applications such as signal timing adjustments, vehicle advisory messages, and incident notifications. The proposed DT is a vital tool in T-CPS, enabling real-time traffic monitoring, prediction, and optimization to enhance the safety and mobility of transportation systems.

Given the current state of vehicle automation, understanding the similarities and differences between Automated Vehicle (AV) and Human-driven Vehicle (HDV) crashes is crucial to identifying specific challenges and areas for improvement of AV technology. The challenges of directly comparing AV and HDV crashes include differing traffic environments, crash reporting discrepancies, and underreporting of HDV crashes. To address these challenges 555 AV crashes and 39,270 HDV crashes are used to perform a Crash Sequence Analysis. Results show that while AV and HDV crashes can be classified into similar groups based on vehicle movements, the types of crashes can differ significantly. While intersections pose greater challenges for AVs compared to HDVs, the severity of AV crashes is lower, which might be attributed to the reduced number of crashes involving AVs and vulnerable road users. Considering similar crash contexts, AV and HDV crashes can differ significantly, especially in scenarios involving pedestrians, left and right turns, the stopping movement of AVs, and red light violations. Furthermore, the analysis shows three groups with contexts unique to the AV crashes (rear end crashes following a lane change or stopped vehicles, and side swipe crashes on narrow streets), which can indicate both technological challenges and the differences in crash exposure caused by the manufacturers' training environment. The above emphasizes the importance of addressing expectancy violations likely to emerge in a mixed fleet environment of AVs and HDVs, particularly accounting for geographical and cultural specificities. These findings provide key insights for shaping future AV development, guiding public sector decisions, and building public trust in automation.

Level 4 automated vehicles (AVs) with the operational design domain (ODD) expanding over time are expected to be the future. Although Level 4 AVs do not require driver takeover, human driving will be necessary outside the ODD. While there is a significant amount of research on takeover/disengagement, no prior studies have explored the safety challenges of manual operation of Level 4 AVs. Crash sequence analysis was employed to compare crashes of the AV (during manual control) (AVM) and general driving population, using U.S. data from California Department of Motor Vehicles crash reports and the Crash Report Sampling System (CRSS) dataset, respectively. Clusters of AVM and CRSS crashes were aggregated into nine groups based on crash context. The results suggest that certain crash groups are more challenging for AVM than for CRSS. AVM crashes are vastly less severe than CRSS crashes for all but one crash group that involved right turns. Nearly half of the AVM crashes involving left and right turns were rear-end crashes, while the majority of similar CRSS crashes were side-swipe or angle. The majority of rear-end AVM crashes occur at intersections, while the converse is true for similar CRSS crashes. Intriguingly, in all the AVM rear-end crashes, the lead vehicle was an AV, suggesting hesitation on the part of the safety driver. For AVM, while lane-changing crashes were less frequent, crashes involving parked vehicles were more frequent than for CRSS. The findings indicate the importance of understanding how driver behavior changes with Level 4 AVs, and how driver training might play an important role in the safety of AVs.

The advancement of Connected and Automated Vehicles (CAVs) and Vehicle-to-Everything (V2X) offers significant potential for enhancing transportation safety, mobility, and sustainability. However, the integration and analysis of the diverse and voluminous V2X data, including Basic Safety Messages (BSMs) and Signal Phase and Timing (SPaT) data, present substantial challenges, especially on Connected Vehicle Corridors. These challenges include managing large data volumes, ensuring real-time data integration, and understanding complex traffic scenarios. Although these projects have developed an advanced CAV data pipeline that enables real-time communication between vehicles, infrastructure, and other road users for managing connected vehicle and roadside unit (RSU) data, significant hurdles in data comprehension and real-time scenario analysis and reasoning persist. To address these issues, we introduce the V2X-LLM framework, a novel enhancement to the existing CV data pipeline. V2X-LLM leverages Large Language Models (LLMs) to improve the understanding and real-time analysis of V2X data. The framework includes four key tasks: Scenario Explanation, offering detailed narratives of traffic conditions; V2X Data Description, detailing vehicle and infrastructure statuses; State Prediction, forecasting future traffic states; and Navigation Advisory, providing optimized routing instructions. By integrating LLM-driven reasoning with V2X data within the data pipeline, the V2X-LLM framework offers real-time feedback and decision support for traffic management. This integration enhances the accuracy of traffic analysis, safety, and traffic optimization. Demonstrations in a real-world urban corridor highlight the framework's potential to advance intelligent transportation systems.

Performance-based design is a renewed approach for project decision-making to specifically address the purpose and needs of projects by providing design flexibility. Resources to evaluate roadway design based on safety performance are an important need for transportation practitioners. In this study, naturalistic driving study (NDS) data were used to estimate the safety effect of elements that influence driving behavior on rural undivided two-lane horizontal curves. Available data included 3,292 horizontal curves and 150,233 traversals, which required a significant data processing effort to conduct data cleaning and quality assessment. From the safety surrogates evaluated and methods implemented, lane position provided the most consistent and statistically significant results. Centerline and edge line encroachment events were modeled with the negative binomial using traffic volume and curve geometry as the predictor components. Encroachment estimates were associated with observed crashes from state data to convert encroachments to crashes. Predictor variables such as curve radius showed a decreasing trend in predicted crashes as curve radius increased. Similarly, as shoulder or lane width increased, predicted crashes decreased. Crash estimates derived from safety surrogates were used to develop an analytical tool aimed at practitioners for curve design considerations. Data input includes curve radius, shoulder width, lane width, curve length, traffic volume, expected service life, and construction cost. The economic assessment provides a quantitative measure for practitioners to evaluate alternative curve designs by assessing the tradeoffs between safety and costs of implementation.

An accurate and robust localization system is crucial for autonomous vehicles (AVs) to enable safe driving in urban scenes. While existing global navigation satellite system (GNSS)-based methods are effective at locating vehicles in open-sky regions, achieving high-accuracy positioning in urban canyons such as lower layers of multi-layer bridges, streets beside tall buildings, tunnels, etc., remains a challenge. In this paper, we investigate the potential of cellular-vehicle-to-everything (C-V2X) wireless communications in improving the localization performance of AVs under GNSS-denied environments. Specifically, we propose the first roadside unit (RSU)-based cooperative localization framework, namely CV2X-LOCA, that only uses C-V2X channel state information to achieve lane-level positioning accuracy. CV2X-LOCA consists of four key parts: data processing module, coarse positioning module, environment parameter correcting module, and vehicle trajectory filtering module. These modules jointly handle challenges present in dynamic C-V2X networks. Extensive simulation and field experiments show that CV2X-LOCA achieves state-of-the-art performance for vehicle localization even under noisy conditions with high-speed movement and sparse RSU coverage environments. While focusing on AV localization, CV2X-LOCA also can extend to other C-V2X-equipped road users. The study results also provide insights into future investment decisions for transportation agencies regarding deploying RSUs cost-effectively.

Cellular vehicle-to-everything (C-V2X) is an emerging technology that has the potential to significantly improve traffic safety and efficiency through various applications such as pedestrian alerts, traffic hazard warnings, signal timing information broadcasting, etc. With the help of the Federal Communications Commission's agreement to waive restrictive rules on C-V2X deployment, extensive real-world testing and implementation of C-V2X is on the horizon. However, the performance of C-V2X in urban environments remains unclear as existing studies have been focusing on evaluating it in either simulated or highway environments. To facilitate the understanding of its performance in urban environments, this study has conducted a systematic evaluation of the efficiency and reliability of C-V2X using various metrics based on real-world data collected through roadside units (RSUs) and onboard units in urban environments. Experimental results indicate that C-V2X has relatively stable efficiency (i.e., latency) under different conditions. However, its reliability (i.e., packet error rate and spatial delivery rate) is significantly affected by multiple factors, including the distance from the vehicle to the RSU and to its surrounding buildings. Moreover, this study has proposed a concept of machine learning-assisted RSU planning to facilitate the planning and deployment of C-V2X in urban areas. An XGBoost model is developed to accurately estimate the probability of the delivery of V2X messages using C-V2X. Results from explaining the model suggest that the planning of RSUs should consider the spacing of RSUs and the density of buildings in the area.

In the United States, 91.4 cm (3.0 ft) has been considered a safe lateral distance when passing bicyclists; however, this threshold may be more suitable to urban environments where most bicycle safety research has originated—limited research is available on rural roadways. In this study, a road bicycle was instrumented to collect naturalistic data and analyze vehicle-bicycle lateral passing distances (LPD) on high-speed rural two-lane roadways with paved shoulders. The bicycle was ridden in the center of the shoulder. Overall, 2,119 observations of vehicles passing the instrumented bicycle were obtained from a study segment. The impacts of different variables on distributions of vehicle-bicycle LPD measurements were analyzed. Passing vehicles with opposing traffic passed the bicyclist significantly closer than passing vehicles without opposing traffic. More predominant with opposing traffic, heavy vehicles passed the bicyclist closer than passenger vehicles. Similarly, following vehicles passed the bicyclist closer than leading vehicles. The study segment was repaved during the data collection period, so data before and after the roadway improvement was available. Vehicles with old pavement/marking passed the bicyclist closer than vehicles with new pavement/marking conditions. Distribution of LPD measurements was skewed to the left towards smaller values and had a long tail to the right towards larger values. Through maximum likelihood estimation, the Log-normal distribution was found to best fit the data distribution. Using empirical and theoretical density distributions, LPD was formally specified as a function of opposing traffic with a theoretical mixture distribution and generalized to estimate the probability of vehicles passing bicyclists below a certain threshold.

Safety is one of the most essential considerations when evaluating the performance of autonomous vehicles (AVs). Real-world AV data, including trajectory, detection, and crash data, are becoming increasingly popular as they provide possibilities for a realistic evaluation of AVs’ performance. While substantial research was conducted to estimate general crash patterns utilizing structured AV crash data, a comprehensive exploration of AV crash narratives remains limited. These narratives contain latent information about AV crashes that can further the understanding of AV safety. Therefore, this study utilizes the Structural Topic Model (STM), a natural language processing technique, to extract latent topics from unstructured AV crash narratives while incorporating crash metadata (i.e., the severity and year of crashes). In total, 15 topics are identified and are further divided into behavior-related, party-related, location-related, and general topics. Using these topics, AV crashes can be systematically described and clustered. Results from the STM suggest that AVs’ abilities to interact with vulnerable road users (VRUs) and react to lane-change behavior need to be further improved. Moreover, an XGBoost model is developed to investigate the relationships between the topics and crash severity. The model significantly outperforms existing studies in terms of accuracy, suggesting that the extracted topics are closely related to crash severity. Results from interpreting the model indicate that topics containing information about crash severity and VRUs have significant impacts on the model’s output, which are suggested to be included in future AV crash reporting.

As automated vehicles (AVs) gradually gain prevalence on public roads, understanding their distinctive driving behavior is crucial for traffic management and planning. This study conducted field experiments using an SAE Level-3/4 AV and collected driving data of AVs and human drivers on public roads using sensors including GPS, radar, camera, and LiDAR. The Wilcoxon rank-sum test is used to identify the difference in the behavior between AVs and human drivers. In addition, logistic regression and Extreme Gradient Boosting (XGBoost) are used to classify AVs and human drivers. Results suggest that there exists a significant difference in driving behavior between AVs and human drivers. Moreover, features including the mean speed and the distance from the vehicle to the detected objects are positively related to the probability of the vehicles being AVs, while the standard deviation of speed and the mean acceleration are negatively associated with it. Furthermore, XGBoost accurately identifies AVs and human drivers using the extracted features with an average area under the curve of 0.92. Results from interpreting results from XGBoost indicate that it performs better when the mean speed is either in the low or high ranges. Moreover, AVs and human drivers are hard to differentiate using the model when the vehicle is too far from other objects. This study underscores the substantial divergence in driving behavior between AVs and human drivers, offering valuable insights for the evaluation of the impact of AVs on traffic conditions.

The flashing yellow arrow (FYA) indication has become commonly used for communicating permissive left-turn operations to road users. A conventional signal phasing communicated to left-turning drivers consists of a leading protected phase followed by a permissive phase using a FYA indication, often referred to as protected-permissive. To transition from the protected to the permissive phase, change (steady yellow arrow) and clearance (steady red arrow, all-red) intervals, or only a change interval, may be used. There is no specific guidance on the use of clearance intervals for left turns with protected-permissive phase and FYA indication. In this study, field data were collected from different geographical regions in the United States to evaluate change and clearance intervals with the objective of developing guidance. Video data were recorded at 37 intersections during peak hours, for approximately 142 h, across eight states. Overall, 4,001 observations of vehicles turning left during the change or clearance interval were analyzed. Field observations were evaluated at the approach level to assess left-turn violation rate and violations per cycle as a function of change and clearance interval configuration. At the individual vehicle level, logistic regression was implemented to evaluate the effect of change interval and clearance interval duration, delayed onset of FYA, and regional variation. The results of the analysis at the approach and individual vehicle level were consistent and indicate that: (i) a clearance interval should be included, (ii) delaying the onset of FYA indication with an extension beyond the all-red clearance interval would reduce left-turn signal violations, and (iii) duration of change intervals has a marginal impact on left-turn signal violations.

Using video, interactions between right-turning vehicles and pedestrians were documented for sites with a permissive circular green or a flashing yellow arrow (FYA) permissive right-turn indication. Timestamps associated with key vehicle and pedestrian positions were documented and used to calculate three metrics: time for driver to complete right turn (with and without a pedestrian presence) and the time for the pedestrian to arrive at the theoretical conflict point of vehicle–pedestrian interactions. Using timestamp data, multiple non-probabilistic linear regression models were created to describe the relationship between the position of the pedestrian within the crosswalk and the time for a right-turning vehicle maneuver to be completed. Model results implicitly describe the attitude of drivers toward pedestrian presence on the crosswalk. Larger right-turning maneuver completion times are arguably an indicator of greater respect toward pedestrian presence. Given the nature of the model’s output, a pedestrian respect indicator (PRI) based on the value of model parameters is introduced as a potential indicator of the safety of vehicle–pedestrian interactions. PRI values can be computed for individual sites thus allowing its use as a safety indicator. The higher the PRI, the more “respect” toward pedestrians. Proof-of-concept modeling results based on available timestamp data suggest that interactions controlled by a right-turn FYA indication have a higher PRI than interactions controlled by a circular green. An argument is made that PRI values can be potentially used to proactively evaluate the effectiveness of traffic control devices and other safety countermeasures without the need for crash data.

The COVID-19 pandemic had a significant impact on mobility worldwide, specifically through stay-at-home orders. There is a general consensus in the literature regarding the reduction of traffic during that period, and a trend toward increased speeds. However, the literature is still scarce regarding the pandemic's long-term and site-specific effects on traffic volumes and speed behavior. In this context, the present study looks at speed and traffic volume data, using several temporal and spatial filters to isolate the effect of the pandemic from other elements that can influence speeds, such as congestion, construction, weather, and faulty detector data. Data from Wisconsin for the stay-at-home period in 2020 and corresponding time periods in 2019 and 2021 were used. Traffic volume and speed were analyzed using descriptive and inferential (Kolmogorov–Smirnov Test and Mann Whitney U Test) statistics. While the results for traffic showed the expected reduction in 2020 in relation to other years, speeds also showed a reduction in 2020 for eight of the twelve analyzed locations, most of these rural areas. Furthermore, the results show a return of speeds to pre-pandemic levels in 2021, associated with the partial or complete recovery of traffic volumes.