High Traffic Volume Research Articles

Cyclists' perception of cycling infrastructure – A Repertory Grid approach

Previous research indicate that cyclists prefer safe and comfortable infrastructure. However, concepts like safety or comfort may be understood and defined in different ways by different individuals. Moreover, these concepts are so broad that it is unclear to which specific characteristics of the traffic infrastructure they are linked. Therefore, the aim of this study is to inductively examine individual evaluation criteria that cyclists use to perceive and evaluate certain route attributes. Using the Repertory Grid technique, we elicited the personal constructs of 23 participants about given route attributes and asked them to rate every attribute on every construct. The constructs were categorized and clustered resulting in five evaluation criteria, namely Mental Comfort, Physical Comfort, Interaction, Environment, and Ease of Use. Both comfort clusters were rated the most relevant for route choice, but they strengthen the suggestion of two distinct aspects of comfort, one referring to mental stress, the other referring to physical effort. Furthermore, the analyses revealed Interaction as a relatively new route criterion which also raises a new view on the negative evaluation of motor traffic and expands the concept of stress. Regarding the evaluation of route attributes, a high traffic volume and cobbled stone were rated the most negative, whereas separated cycling facilities were evaluated as most positive. Findings of this study expand existing research by qualitative insights and provide a more detailed understanding of route criteria like comfort or safety. This can be used to enhance cycling facilities and to offer preferable infrastructure for cyclists.

Rockfall Vulnerability of a Rural Road Network—A Methodological Approach in the Harz Mountains, Germany

Mass movements are linked to increasing amounts of damage and disruptions to transportation infrastructures. A valid risk assessment in order to reduce future costs is not always appropriate, as adequate information on landslide data is missing. The presented study estimates the rockfall susceptibility on a rural road network in the Harz mountains using a bivariate statistical method (information value method). The model is validated using a receiver operating characteristic (ROC) analysis. In addition, the vulnerability of the road network is estimated using vulnerability indicators. The susceptibility model assigns a high or very high susceptibility to 23% of the area in the road network corridor. The relevant road sections are linked to high slope values, NE orientations of road sections, and low-to-moderate vulnerability values. The highest vulnerability values can be found on marginal road sections with high average daily traffic volumes. The combination of the presented methods proposes an easily applicable estimate of vulnerability where conventional methods (i.e., vulnerability curves, matrices) cannot be implemented.

Behavior Model Analysis of Pedestrians and Cyclists Mode Selection Base on Active Transportation on Universitas Jember Tegalboto Campus

Abstract The University of Jember as one of the campuses that have a high traffic volume is trying to realize a “Green Campus” through the “Ring the Bell” program. However, this vision has not been implemented optimally because there are still many campus residents who do not use active modes of transportation (walking & cycling) even though sidewalk facilities are available. This indicates that the behavioral preferences for choosing active transportation modes for the residents of the Jember University campus are still unclear. Therefore, this study aims to analyze the behavioral model of pedestrian and cyclists mode selection based on active transportation at the Tegalboto Campus, Jember University. The primary data collection of residents’ preferences was obtained through the TPB (Theory Of Planned Behavior) questionnaire. SEM model (Structural Equation Model) was applied in the behavioral analysis process to determine the interaction between the independent variables of TPB as a whole. The results of behavioral modeling show that intention is the variable that most influences the behavior of campus residents in using active transportation modes, while attitude is the variable that most influences intentions than subjective norms and perceptions of behavioral control.

Hotspot identification on urban arterials at the meso level

Urban arterials form the main structure of city street networks, and typically have high traffic volume and high crash frequency. To reduce the number of crashes, hotspot identification (HSID) is the first step in the traffic safety management process, and often utilizes crash prediction models. Classical crash prediction models investigate the relationship between roadway characteristics and traffic safety at the micro level, that is, they treat road segments and intersections as isolated units. Micro-level analysis has limitations, however, when examining urban arterial crashes: 1) signal spacing is typically short for urban arterials in dense street networks, and there are interactions between intersections and road segments that classical models do not accommodate; and 2) for practical engineering, a hotspot to which countermeasures are applied generally consists of several adjacent intersections and road segments instead of a single intersection or road segment. To address these concerns, signalized intersections and their adjacent road segments were combined into meso-level units, which were adopted to investigate traffic safety data from 21 urban arterials in Shanghai, China. To determine if the meso-level unit is the most suitable research unit for identifying hotspots on urban arterials, and if so, which HSID method can most consistently identify them, this study identified micro-level (separate intersection and road segment) and meso-level (combined intersection and road segment) hotspots using crash frequency, empirical Bayesian (EB), potential for safety improvement (PSI), and full Bayesian (FB) methods. To evaluate the performance of the HSID methods, hotspot consistency over two years was tested. The results showed that 1) EB and PSI performed better than the other methods no matter which research unit was used; 2) substantial inconsistency between the identified micro- and meso-level hotspots. To identify both hazardous corridors and individual intersections and road segments on urban arterials, a combination of micro- and meso-level hotspots should be recommended to local transportation authorities.

An integrated air quality modeling system coupling regional-urban and street models in Beijing

Recently, ozone (O3) pollution has become a significant problem for Beijing owing to its high traffic volume. To improve the simulation of street-scale nitrogen oxides (NOx) and O3, an integrated air quality modeling system coupling regional urban/street (IAQMS-street) was developed for Beijing. A weather research and forecasting atmospheric model, regional nested air quality prediction system, model of urban network of intersection canyons and highways, and a real-time on-road emission model were coupled by a downscaling method. A two-week simulation with high O3 concentrations (September 25, 2020 to October 8, 2020) was conducted to evaluate the impact of high-resolution traffic emissions and the street-scale simulation by three sensitivity scenarios. The results showed that IAQMS-street improved the model performance for NO, NO2, and O3 compared to regional models. The correlation coefficients (R) and normal mean bias (NMB) increased from ~0.3 to ~0.6 and from −83.3% to 14.8% for NO, respectively. For O3, the R increased by ~10%, and the NMB decreased from 48.7% to −4.4%. In particular, the model performance for O3, NO, and NO2 in high-concentration periods was significantly improved by IAQMS-street. Both high-resolution vehicle emissions and street-scale simulations contributed to this improvement. Our results showed that IAQMS-street provides a more realistic tool for simulating and resolving finer urban photochemical pollution heterogeneity in Beijing.

A Software-Defined Programmable Testbed for Beyond 5G Optical-Wireless Experimentation at City-Scale

Wireless traffic will significantly increase in next-generation networks. Therefore, there is a need for novel optical network front-/mid-/backhaul (x-haul) architectures that can deliver ultra-high bandwidth and ultra-low-latency services to wireless and edge cloud systems, as well as support various cloud radio access network (C-RAN) architectures. Designing and evaluating such architectures requires large-scale experimental platforms. In this article, we present the design and implementation of a unique, remotely accessible, disaggregated, and programmable optical network as the key component of the city-scale PAWR COSMOS advanced wireless testbed. We discuss the design and implementation of a dedicated multi-functional Ryu software-defined networking (SDN) controller for the testbed's optical network wavelength channel assignment and topology reconfiguration. We present an example use of the SDN controller for monitoring and automated measurement of the wavelength-dependent amplifier gain in deployed reconfigurable optical add/drop mul-tiplexer (ROADM) units. We also demonstrate a pilot experiment focusing on a wireless handover scheme for C-RAN architectures in the COSMOS testbed, where high traffic volume can be supported by dynamic wavelength allocation via optical switching in the fronthaul. The field-deployed testbed and SDN controller can be used by the research community to push the envelope in the area of optical networks for beyond 5G and 6G.

Concentration and leachability of N-(1,3-dimethylbutyl)-N′-phenyl-p-phenylenediamine (6PPD) and its quinone transformation product (6PPD-Q) in road dust collected in Tokyo, Japan

A recently identified chemical, 2-((4-Methylpentan-2-yl)amino)-5-(phenylamino)cyclohexa-2,5-diene-1,4-dione (6PPD-quinone; 6PPD-Q), is a transformation product of an additive used in the manufacture of tire rubber and causes acute lethality in coho salmon (Oncorhynchus kisutch) in urban watersheds. Despite its potential presence and ecotoxicity in receiving waters worldwide, information on the occurrence and fate of 6PPD-Q is limited. Here, we investigated the concentrations of 6PPD-Q and its parent chemical, 6PPD, in road dust collected from arterial and residential roads in Tokyo, Japan from May to October 2021. 6PPD-Q concentrations were highest from May to June, when atmospheric ozone concentrations are the highest in Japan; a correlation between 6PPD-Q and photochemical oxidants, as an alternative to ozone, corroborated this finding. We also found that 6PPD-Q concentrations at photochemical oxidant concentrations ranging from 35 to 47 ppbv were higher in dust collected from roads with high traffic volumes (i.e., arterial roads; median: 8.6 μg/g-OC) than in dust collected from roads with lower traffic volumes (i.e., residential roads; median: 6.3 μg/g-OC), indicating that 6PPD-Q is generated from traffic-related sources. We also found that 6PPD-Q was leached from dust particles within a few hours, with a log partitioning coefficient between organic carbon and water (KOC) of 3.2–3.5. The present results will help to understand the environmental occurrence, fate, and behavior of 6PPD-Q.

Flow-level coordination of connected and autonomous vehicles in multilane freeway ramp merging areas

On-ramp merging areas are deemed to be typical bottlenecks for freeway networks due to the intensive disturbances induced by the frequent merging, weaving, and lane-changing behaviors. The Connected and Autonomous Vehicles (CAVs), benefited from their capabilities of real-time communication and precise motion control, hold an opportunity to promote ramp merging operation through enhanced cooperation. The existing CAV cooperation strategies are mainly designed for single-lane freeways, although multilane configurations are more prevailing in the real-world. In this paper, we present a flow-level CAV coordination strategy to facilitate merging operation in multilane freeways. The coordination integrates lane-change rules between mainstream lanes, proactive creation of large merging gaps, and platooning of ramp vehicles for enhanced benefits in traffic flow stability and efficiency. The strategy is formulated under an optimization framework, where the optimal control plan is determined based on real-time traffic conditions. The impacts of tunable model parameters on the produced control plan are discussed in detail. The efficiency of the proposed multilane strategy is demonstrated in a micro-simulation environment. The results show that the coordination can substantially improve the overall ramp merging efficiency and prevent recurrent traffic congestions, especially under high traffic volume conditions.

Objective and Perceived Traffic Safety for Children: A Systematic Literature Review of Traffic and Built Environment Characteristics Related to Safe Travel.

The literature on children’s active transportation has shown the influence of the built environment characteristics on walking and crashes. Various reviews have examined those two questions. One influence on walking is the perception of traffic safety. However, it is not clear how, or even if, the built environment affects such perceptions. This research aims to understand which traffic and built environment characteristics influence objective and subjective/perceived traffic safety for children based on the analysis of previous studies in the field. Two types of research were used: the first examines the association between traffic and built environment characteristics and child pedestrian and/or cyclist collisions/injuries; the second relates to the perception of safety by parents and children for active transportation and, where studied, its relationship with built environment characteristics. A systematic review was conducted using five electronic databases. The total number of articles retrieved was reduced to 38 following the eligibility criteria and quality assessment, where 25 articles relate to injuries among children and 13 articles pertain to perception of safety. The results showed that high traffic volume and high vehicle speed are the main reasons children and parents feel unsafe when children use active travel, which matches the main findings on objective safety. Few articles on perception of safety related to the objective built environment were found. However, consistent findings exist. The presence of sidewalk was related to the safety of children. The presence of a crossing guard was positively related to perceived safety but was associated with higher rates of injuries among children. Intersection density was related to unsafe perceptions but was not statistically associated with objective traffic safety. Additionally, population density was found to be positively related to injuries among children, but not to perception of safety. The results help policy strategy to enhance the safety of children when using active transport modes.

Competence of Bus Rapid Transit Systems Coupled with Transit Signal Priority at Signalized Junctions

Abstract One of the primary causes of poor public transport performance is delays at intersections. Among the efficient and sustainable solutions to boost mass transportation performance, Bus Rapid Transit (BRT) consists of infrastructures integrating dedicated bus lanes and smart operational service with different ITS technologies like Transit Signal Priority (TSP). This research studies the competence of buses operating on junctions of the BRT corridor where they have Signal Priority on the dedicated lane. The studied intersection is located around the center of the Addis Ababa BRT-B2 line, which is relatively gentle grade and characterized by the high traffic and pedestrian volume. Microscopic models were created for the chosen intersection, along with possible calibration and validation; moreover, a statistical comparison was performed to evaluate different scenarios with the goal of displaying the deployment benefits. To assess the performance of BRT buses and their overall influence on general traffic, scenarios with and without TSP were evaluated. PTV VISSIM and the VisVAP add-on simulation program were used to examine TSP alternatives. Incorporating TSP reduced the travel time by up to 4.78% in the priority direction, the average travel speed increased by 7.25%, and the queue length also reduced by a maximum of 6%, whereas in the non-priority direction, the queue length increased by a maximum of 2.5%. Moreover, the overall average passenger delay has reduced by an average amount of 15%. One of the simplest ways to improve transit performance could be signal priority strategies, which has a minor influence on the general traffic.

Performance Analysis of Road Section and Unsignalized Intersections in Order to Prevent Traffic Jams on Jl H. Djole – Jl. Pasar Lama

The unsignalized intersection of Jalan H Djole – Jalan Pasar Lama is located in Bekasi City is a commercial area with densely populated settlements and dense economic activity every day. Lack of facilities and infrastructure led to traffic jams on the intersection. A analysis is needed that aims to determine the volume of the vehicle, the performance of road sections and intersections which are measured based on capacity, degree of saturation, and level of service. And then alternative solutions to improve the performance of these roads and intersections.
 The analytical method used is the Indonesian Road Capacity Manual 1997. The data required in this study are primary data in the form of road geometric data, traffic volume, vehicle speed data and side friction data. As well as secondary data such as location maps, land use, and population data. Based on the results of the analysis that has been carried out, the highest traffic volume at the intersection occurs on Friday, 17.00 - 18.00 WIB, with a DS value of 1.06 and LOS F. Alternative solutions that can be given at the intersection are vehicles on minor roads A and C You can only turn left, with the alternative solution, the DS values ​​are 0.73 and LOS C. Whereas on the roads it has DS values ​​of 0.71 and LOS C. An alternative solution that can be given to roads is to reduce the level of side friction around the segment. Street. With this alternative solution, the DS value is 0.65 and LOS B.

Анализ эффективности способов приоритета автобусам при проезде перекрестков с применением адаптивного управления светофорами

In recent years in the cities of the Russian Federation, within the framework of the national project “Safe and high-quality roads” the concept of “priority development of public transport” has been actively implemented and Intelligent transport systems have been created. The article discusses ways to prioritize public transport when passing through intersections using adaptive traffic light control as a possible alternative to a lane for route vehicles. The aim of the study is to reduce the loss of time for road users by determining the best ways to prioritize buses when passing regulated intersections, taking into account the number of passengers in public transport and the loss of time for all traffic participants. During the study, field observation methods, mathematical and simulation modeling were used. The account of the moments of the approach of buses to the regulated intersection and the analysis of the loss of time for buses when passing through the section on the central street of Tyumen was carried out. The scientific novelty lies in establishing the dependencies of the average speed of the bus on the approach to the intersection from the moment until the end of the traffic signal, the delay time of the buses on the number and moment of stopping the buses when passing the intersection. With the help of simulation micromodeling, the dependence of the delay time of buses and cars on traffic intensity was established for five options for priority to public transport. Two variants of priority for buses when passing through intersections are based on the methods of adaptive control of traffic lights and the algorithms “stage recall” and “green extension”. The use of a dedicated lane and the “stage recall” method at high traffic intensity reduces the delay time for buses by 94%, for personal vehicles – increases by 85%. At low traffic intensity, the delay time changes by 11 and 6%, respectively, for buses and cars. The use of the “green extension” method allows to reduce the total delay time by 3.2% with a high traffic volume (1900 veh/h) and a small number of buses (40 veh/h) and has a greater effect than the lane for fixed-route vehicles (deterioration by 148%). To determine the most efficient priority method, it is proposed to take into account the number of passengers in public transport. The dependence of the total delay time of all traffic participants on the traffic intensity of cars and buses with different priority methods has been established. This approach allows us to further determine the area of rational application of priority methods for public transport with the target function of reducing the loss of time for all traffic participants.

A fog-based approach for distributed cloud testing: a VR-assist rental testing case study

Cloud computing offers the opportunity to improve productivity and reduce costs. However, this technology faces high latency problems. In this context, a decentralised cloud paradigm has been introduced in order to bring geographically cloud services close to users. Moreover, the growth of personal devices has stimulated cloud services with further requirements such as compliance (conformance) with a standard or high traffic volume. To cope with these problems, we need assessments to test for example how these requirements can affect the performance degradation in such environments and how we can enhance the distributed cloud architecture in order to perform the end-users objectives. The main contribution in this paper is to suggest a distributed testing approach derived from 'fog computing' design and leveraging its advantages especially optimising response time requirement. To this end, we present a comprehensive survey on distributed cloud architectures, and then we formally suggest a new architecture for testing in such environments. Finally, we illustrate our approach through a case study.

Determination of polycyclic aromatic hydrocarbons extracted from lichens by gas chromatography–mass spectrometry

Lichens are well-known biomonitors for semi-volatile pollutants, due to their ability to absorb and retain different chemical compounds such as Polycyclic Aromatic Hydrocarbons (PAHs), directly linked to levels in the atmosphere. Based on that, this paper proposes an analytical method capable of quantifying 16 EPA-PAHs from lichens found in an intertropical zone, as a natural alternative to typical capture methods, with the aim of monitoring atmospheres polluted by toxic compounds. An analytical protocol, including sample pre-treatment, followed by ultrasound extraction, clean-up in a chromatographic column, concentration and quantification by Gas Chromatography-Mass Spectrometry (GC-MS) using Selective Ion Monitoring has been developed. Additionally, a set of guidelines on lichen collection and sample handling is given, in order to achieve representative samples.•Limits of quantification (LOQ) and detection (LOD) varied from 2.0 to 16 µg/L and 1.0 to 5.0 µg/L, respectively. Calibration curves had correlation coefficients higher than 0.99 in all cases.•Validation of the method for determining PAHs concentration associated to 30 lichen samples collected along two roads, with high and low traffic volumes was carried out.•The method showed good performance according to the sources of PAHs, traffic patterns and gradient in roads.

Application of GIS to assess the accessibility of urban streets in Baghdad city

Accessibility has an important impact on shaping human activity patterns on all of the spatial scales. This study presented an evaluation of accessibility levels with private to commercial centers for three selected routes in Baghdad city. The study involved more than 45 days transport survey for private vehicles in Baghdad city using Global Positioning System (GPS) probe for recording indicators of traffic performance. Gravity model was used to measure accessibility index as an implementation of GIS-based model by using link geography and the spatial boundary of analysis in order to build route networks at three routes in Baghdad City, Bayaa intersection - Bab Al-Muatham intersection (Route 1), Bayaa intersection - Bab Al-Muatham intersection (Route 2) and 14 Ramadan Street - Bab Al-Muatham intersection (Route 3). It was found that Route 1 has the high accessibility index with 0.67 in compare with Route 2 and 3 (0.58 and 0.59), respectively. The reason that Route 1 had the highest accessibility index due to the high access point and low traffic volume as compared with the other two routes.

Efficiency Assessment of a Signalized Roundabout and a Traffic Intersection in Baghdad City

In Baghdad city, Iraq, the traffic volumes have rapidly grown during the last 15 years. Road networks need to reevaluate and decide if they are operating properly or not regarding the increase in the number of vehicles. Al-Jadriyah intersection (a four-leg signalized intersection) and Kamal Junblat Square (a multi-lane roundabout), which are two important intersections in Baghdad city with high traffic volumes, were selected to be reevaluated by the SIDRA package in this research. Traffic volume and vehicle movement data were abstracted from videotapes by the Smart Traffic Analyzer (STA) Software. The performance measures include delay and LOS. The analysis results by SIDRA Intersection 8.0.1 show that the performance of the roundabout is better than the signalized intersection but experiences high delay, and low LOS. Therefore, alternatives are proposed to improve the performance for current and future traffic volumes with low-medium delays.

On Site Improvement of Fines-Rich Unbound Granular Materials with Hydrophobic Polymer and Lime

Many roads that were initially designed for relatively low traffic volumes need re-surfacing or partial replacement of the unbound granular material to satisfy current traffic demand. Significant research efforts based on laboratory studies have been seen in the literature to characterize the suitability of virgin materials, which is relatively expensive and unsustainable. Therefore, the object of this study is the in situ recycling of existing materials in two road sections by improving their properties with a suitable additive. A hydrophobic synthetic polymer was chosen for two trials due to the high plasticity of fines of the in situ materials and a high chance of water intrusion in the low-lying plains in Adelaide. The extensive laboratory characterization shows that hydrophobicity is imparted in capillary rise tests, improved drainage in permeability tests, and greater matric suction at the same moisture content. Furthermore, the unconfined compressive strength was increased. The repeated loading triaxial testing showed higher stiffness and lowered permanent strain to withstand higher traffic volume. In general, in situ recycling is adaptable and considered to be cheaper and sustainable. The estimated current costs and carbon footprints are presented for re-construction and in situ recycling with dry powder polymer, or solely with lime, to help construction planning.

Inferring the causal effect of work zones on crashes: Methodology and a case study

The increasing number of crashes occurring in work zones has received considerable attention in recent years. Previous studies have mainly focused on associations between work zone configurations and crash occurrence. Although identification of associational relations helps us understand how work zones co-exist with crashes, it does not provide interventional guidelines necessary to improve safety of work zone operations. In this paper, a causal inference model based on the potential outcome framework is proposed to rigorously infer the causal effects of work zone presence on crash risks under various work zone configurations, along with robustness tests. In developing such a causal model, three research gaps are identified and addressed: (1) potential confounding bias due to unobservable roadway characteristics; (2) potential bias caused by unobserved variables in multisource data; and (3) lack of actually observed traffic data and weather information at the exact time when a crash occurred and lack of large-scale high-granular data. The proposed methodology is applied to 5,006 work zones in Pennsylvania from 2015 to 2017, and the results are validated via a series of robustness tests. The results show that the causal effect of a work zone on crash occurrence is significantly positive, especially on roadways with high traffic volumes, on long-distance work zones, and work zones conducted during daytime. It appears that conducting work zones during nighttime with the current deployment strategies on Pennsylvania state roads does not necessarily increase crash risks, but a work zone significantly increases crash risks during day time.

Evaluation of Rapid Repair of Concrete Pavements Using Precast Concrete Technology: A Sustainable and Cost-Effective Solution

Abstract Concrete and asphalt are the two competitive materials for a highway. In Sweden, the predominant material for the highway system is asphalt. But under certain conditions, concrete pavements are competitive alternatives. For example, concrete pavements are suitable for high-traffic volume roads, roads in tunnels, concentrated loads (e.g., bus stops and industrial pavement). Besides the load-carrying capacity, the concrete pavement has many advantages such as durability (wear resistance), resistance against frost heave, environment (pollution, recycling, and low rolling resistance leading to fuel savings), fire resistance, noise limitations, brightness, evenness and aesthetics. Concrete pavements are long-lasting but need final repair. Single slabs may crack in the jointed concrete pavement due to various structural and non-structural factors. Repair and maintenance operations are, therefore, necessary to increase the service life of the structures. To avoid extended lane closures, prevent traffic congestions, and expedite the pavement construction process, precast concrete technology is a recent innovative construction method that can meet the requirement of rapid construction and rehabilitation of the pavement. This paper evaluates rapid repair techniques of concrete pavement using precast concrete technology by analysing three case studies on jointed precast concrete pavements. The study showed that the required amount of time to re-open the pavement to traffic is dramatically reduced with jointed precast concrete panels.

Estimating heterogeneous treatment effects in road safety analysis using generalized random forests

Numerous evaluation studies have been conducted on a variety of road safety measures. However, the issue of treatment heterogeneity, defined as the variation in treatment effects, has rarely been investigated before. This paper contributes to the literature by introducing generalized random forests (GRF) for estimation of heterogeneous treatment effects (HTEs) in road safety analysis. GRF has high functional flexibility and is able to search for complex treatment heterogeneity. We first perform a series of simulation experiments to compare GRF with three causal methods that have been used in road safety studies, i.e., outcome regression method, propensity score method, and doubly robust estimation method. The simulation results suggest that GRF is superior to these three methods in terms of model specification, especially with the existence of nonlinearity and nonadditivity. On the other hand, a large dataset is required for accurate GRF estimation. Then we conduct a case study on the UK’s speed camera program. Our results indicate significant reductions in the number of road accidents at speed camera sites. And the heterogeneity in treatment effects is found to be statistically significant. We further consider the associations between the baseline accident records, traffic volume, local socio-economic characteristics, and the safety effects of speed cameras. In general, the effect of speed cameras is larger at the sites with more baseline accident records, higher traffic volume, and in more densely-populated and deprived areas. Several policy suggestions are provided based on these findings. The evaluation of HTEs likely offers more comprehensive information to local authorities and policy makers, and improves the performance of speed camera programs. Moreover, GRF can be a promising approach for revealing treatment effect heterogeneity in road safety analysis.