Research Review on Traffic Safety for Expressway Maintenance Road Sections

Abstract

With the aging of China’s expressway network, the number of maintenance projects continues to increase, and issues such as construction safety, driving risk, and traffic efficiency have become increasingly prominent. This paper systematically reviews relevant research progress from four aspects: safety characteristics, traffic capacity, work-zone layout, and speed limit management. The review indicates that Western scholars have made extensive use of rich data resources—such as traffic parameters and accident records from expressway maintenance road sections—and have developed relatively systematic and well-established research frameworks in theoretical analysis, practical application, and evaluation methods. In contrast, Chinese studies have mainly relied on specific maintenance projects, commonly employing on-site investigations and traffic simulations to address particular problems, with limited systematization and generalization. Looking forward, it is essential to further strengthen the standardized collection and statistical analysis of traffic data (including accident data) for expressway maintenance road sections. Meanwhile, for complex scenarios such as multi-lane segments, special road sections, reconstruction and expansion sections, as well as extreme climatic conditions and nighttime operations, comprehensive research should be conducted by leveraging new-generation driving simulation, big data analytics, and artificial intelligence technologies, thereby providing scientific support and methodological foundations for building a systematic theoretical framework for traffic safety in expressway maintenance road sections.

1. Introduction

In China, a large number of early-built expressways have gradually entered the mid-to-late stages of their service life, as shown in Figure 1, with their technical conditions continuously deteriorating. Problems such as pavement damage, structural fatigue, and aging of ancillary facilities have become increasingly prominent. As a result, in order to maintain roadway service levels and driving safety, the frequency of maintenance and repair operations have increased year by year, and some sections have had to adopt full or partial lane closures during construction.

China’s expressways are characterized by full closure, high-speed traffic, and high embankments. The combined effects of high-speed traffic flow and complex construction environments expose maintenance road sections to significant traffic safety risks. In response, China has placed great emphasis on traffic safety management during expressway maintenance, successively formulating and implementing a series of technical standards and management specifications, such as Road Traffic Signs and Markings—Part 4: Work Zone (GB 5768.4-2017) [1] and Safety Work Rules for Highway Maintenance (JTG H30-2015) [2]. The introduction of these standards and regulations has effectively improved the standardization and safety of the layout of maintenance road sections, optimized the configuration of traffic safety facilities, and significantly enhanced on-site traffic organization and safety management, thereby providing institutional assurance for the safe and efficient implementation of expressway maintenance operations.

In recent years, the traffic operation environment on China’s expressways has become increasingly complex, characterized by a growing number of lanes, higher driving speeds, an increasing proportion of large vehicles, more frequent lane-changing maneuvers, and increasingly aggressive driving behaviors. These new developments have significantly heightened the uncertainty and dynamics of traffic flow, posing new challenges for traffic safety management and the control of expressway maintenance road sections in the new era.

The traffic safety level of expressway maintenance road sections is not only directly related to the operational safety of construction personnel and mechanical equipment within the work zones but also has a crucial impact on driving safety and traffic order outside the work zones. Western scholars began to pay attention to this field relatively early and have conducted systematic studies based on traffic accident data. As early as 1981, Hargroves [3] performed a statistical analysis of highway traffic accidents in Virginia, USA, from 1977 to 1980, and found that fatal accidents in maintenance road sections accounted for 1.5% of all accidents. The main contributing factors included disorganized traffic management, non-standardized construction operations, vehicle speeding, and drunk driving. Data from the U.S. Fatality Analysis Reporting System (FARS) [4] also showed that, between 1982 and 2017, fatalities in maintenance road sections accounted for 2.05% of all traffic-related deaths, exhibiting an overall upward trend. Pigman et al. [5] further noted that the number of vehicle crashes during maintenance periods was more than 20% higher than during non-maintenance periods. Khattak et al. [6] analyzed crash data from California’s expressway maintenance road sections and found that the rates of property-damage-only and injury crashes increased by 23.8% and 17.3%, respectively. Wang et al. [7] conducted a comprehensive analysis of nighttime expressway maintenance practices in Chongqing city, highlighting that traffic accidents in maintenance road sections are 17.3% more frequent than those in normal sections. The safety hazards within China’s expressway maintenance road sections are equally pressing. As highlighted by Li et al. [8], statistical analysis revealed that vehicle intrusions into work zones were a predominant safety threat, accounting for approximately 70% of all accidents in these areas, primarily due to drivers failing to receive construction information in a timely manner. This underscores that despite the implementation of standards, the effectiveness of on-site information dissemination remains a critical bottleneck, directly contributing to the elevated crash rates observed in both domestic and international studies.

This study aims to conduct a systematic review of the research progress in traffic safety for expressway maintenance road sections, with a particular focus on comparing Chinese and Western scholarly approaches. The review is guided by the following research questions:

(1) What are the key similarities and differences in research focus and methodology between international and Chinese studies?

(2) What are the established findings and remaining challenges in the four key areas of safety characteristics, traffic capacity, work-zone layout, and speed limit management?

(3) What are the promising future research directions that can address the identified gaps?

To address these questions, the paper begins with a detailed account of the systematic literature review methodology in Section 2. It then moves on to Section 3, which presents the findings through an analysis and comparison of research outcomes in the four aforementioned aspects. Finally, Section 4 discusses the implications of these findings, synthesizes the gaps, and proposes a future research agenda.

2. Literature Review Methodology

This paper provides a comprehensive review of research progress on traffic safety in expressway maintenance road sections both in the West and in China. Using two softwares that is CiteSpace 6.3 R1 and Python 3.12, 254 and 284 studies were retrieved from the Web of Science database and the CNKI (China National Knowledge Infrastructure) database, respectively, with the keywords “expressway” and “traffic safety”. A keyword co-occurrence network (shown in Figure 2) and a keyword cloud diagram (shown in Figure 3) were generated by extracting keywords from the retrieved studies. Figure 2 and Figure 3 intuitively present the intellectual structures of studies on expressway traffic safety in Western countries and China. To accurately select the studies required for this paper, the literature selection was conducted through the following two steps. First, potential studies related to the research topic were collected through the aforementioned retrieval. Second, duplicate documents retrieved from multiple channels were removed by manual deduplication, thereby ensuring the uniqueness of the included studies.

The network layout, shown in Figure 2, was optimized using CiteSpace’s built-in force-directed layout algorithm, which spatially groups tightly interconnected keywords. Community detection was performed using the Log-Likelihood Ratio (LLR) algorithm to identify distinct research clusters, which were color-coded. Node size is proportional to the weighted degree centrality, reflecting a keyword’s frequency and co-occurrence strength. This visualization reveals the complex interplay between core themes such as “variable speed limits,” “accident risk,” and “traffic flow characteristics,” underscoring the field’s emphasis on integrated theoretical modeling and risk assessment.

Meanwhile, the keyword cloud diagram, shown in Figure 3, vividly illustrates the major research focuses on traffic safety of expressway maintenance road sections in China. Among these, “traffic safety” appears as the most prominent core keyword, while terms such as “traffic accident,” “traffic engineering,” “traffic simulation,” and “construction area” also stand out with high visibility. These indicate that Chinese research primarily concentrates on accident mechanism analysis, construction safety management, and traffic organization optimization.

From the frequency and co-occurrence strength of the keywords reflected in Figure 2 and Figure 3, the complex interplay between core themes such as safety characteristics, traffic capacity, work-zone layout and speed limit management can be identified. This underscores the emphasis of these research directions on theoretical modeling and risk assessment in relation to traffic safety in this field, which is worthy of particular attention and in-depth discussion. So, through the above comprehensive analysis, this paper’s four primary research themes, which are safety characteristics, traffic capacity, work-zone layout, and speed limit management, were identified.

(1) A systematic analysis of traffic accidents and safety risk assessment in expressway maintenance road sections serves as an essential approach to understanding their safety characteristics. Through statistical examination and mechanism analysis of accident data, the occurrence patterns and underlying influencing mechanisms of traffic accidents can be revealed, and potential risk factors under different stages and types of maintenance conditions can be identified. This process not only deepens the understanding of traffic safety issues in maintenance road sections, but also provides a solid theoretical foundation and decision-making basis for developing scientific and rational safety control strategies, optimizing traffic organization schemes, and enhancing the overall safety management.

(2) During expressway maintenance operations, non-interrupted traffic organization modes are commonly adopted to ensure continuous vehicle passage. These modes typically involve reducing the number of available lanes, closing all lanes on one side, and using the opposite or same-direction temporary lanes for traffic diversion. Although such arrangements help maintain the service function of expressways, they inevitably disrupt the equilibrium of the original traffic flow, causing abrupt changes in traffic characteristics. This substantially increases both the operational risks for maintenance workers and the driving risks for motorists. Moreover, the reduced capacity results in lower vehicle speeds, longer queues, and overall diminished traffic efficiency.

(3) An expressway maintenance road section pertains to the area demarcated between the initial construction warning sign and the construction end sign. Conducting a scientifically rational traffic organization design for the aforementioned areas serves as a crucial measure to alleviate traffic congestion and accidents on maintenance road sections and enhance the service level. Among these aspects, the work-zone layout is of the utmost significance.

(4) Expressway maintenance road sections are typically high-congestion and high-accident areas, functioning as critical bottlenecks that directly affect traffic safety and operational efficiency. Implementing scientifically designed speed limit measures to promote speed convergence within maintenance zones is widely recognized as an essential means of enhancing safety and optimizing traffic organization.

Therefore, this paper systematically selects representative Chinese and Western research achievements from four major dimensions—safety characteristics, traffic capacity, work-zone layout, and speed limit management—and offers an in-depth discussion of their current research progress, limitations, and deficiencies. The findings aim to provide both theoretical and practical guidance for optimizing the layout, traffic organization, and safety management of expressway maintenance road sections, ultimately enhancing traffic efficiency, reducing operational risks, and ensuring driving safety.

3. Research Results and Discussions

3.1. Safety Characteristics of Maintenance Road Sections

3.1.1. Traffic Accident Data Analysis

In Western countries, the statistical system for traffic accidents is relatively well developed, providing solid data support for traffic safety analysis and accident prevention research in expressway maintenance road sections. Garber et al. [9] conducted a statistical analysis of traffic accidents in different areas of maintenance road sections and found that the proportion of accidents occurring in the warning area was significantly higher than in other areas. Pesti et al. [10,11] further examined accident characteristics in expressway maintenance road sections, indicating that over 80% of total accidents occurred in the warning area and upstream transition area, with approximately 60% resulting from forced lane merging due to lane reductions—demonstrating that forced merging behavior was a primary cause of such accidents. Zhao et al. [12] employed mathematical statistical methods to analyze the distribution characteristics of accidents across various control zones, revealing spatial differences in accident occurrence. Khattak et al. [4] developed an analytical model based on multidimensional data—including the number of accidents, severity, average daily traffic volume, location of work zones, and construction duration—and found that accident frequency was significantly and positively correlated with work-zone length, operation time, and traffic volume.

Arditi D et al. [13] investigated day–night safety differences using accident data from Illinois, USA (1996–2001), and found that nighttime construction involved weaker safety awareness among maintenance workers, leading to higher risks at night. In contrast, Akepati et al. [14] reached the opposite conclusion, reporting that drivers tended to be more cautious when passing through maintenance road sections at night, resulting in fewer speeding violations and a lower number of nighttime accidents. Wong et al. [15] analyzed accident data from California expressways (1998–2007) and identified that the most significant factors influencing accident severity were the specific location of the maintenance road section, the time of day when the work was carried out, the duration of construction, and the type of work being performed. Coburn et al. [16] quantitatively analyzed accident types and injury severity, developing an accident cost model to evaluate the economic impacts of traffic incidents. Osman et al. [17] applied an econometric model to investigate collision characteristics between large trucks and passenger cars within maintenance road sections and found substantial differences in the significance and magnitude of influencing factors among different types of maintenance. Park et al. [18] proposed a Lane-Change Risk Index (LCRI) based on fault-tree analysis to evaluate the exposure time and potential collision severity during lane-changing maneuvers, offering a new perspective for studying lane-change safety in maintenance road sections. Sze et al. [19] constructed a multinomial logistic regression model using New Zealand traffic accident data (2008–2013), revealing how roadway environment, driver behavior, and vehicle characteristics influence accident severity. Overall, Western studies have developed a relatively mature research framework encompassing data system construction, model-based analysis, and quantitative risk evaluation, providing a strong theoretical foundation and valuable practical insights for the traffic safety management of maintenance road sections.

Figure 4 presents statistical data on road traffic accidents in China from 2012 to 2019. The data show that, compared with 2012, the total number of traffic accidents in 2019 increased by 16.7%, and the number of fatalities rose by 12.9%, indicating an overall upward trend. Although China’s overall road traffic safety has improved in recent years, the absolute numbers of accidents and fatalities remain high, suggesting that the overall situation is still severe. Due to challenges such as small sample sizes, long observation periods, wide geographical coverage, and low data reliability, the collection of accident data in expressway maintenance road sections is particularly difficult [20]. Consequently, research on traffic accidents in such zones within China remains relatively limited.

A study analyzing traffic accident data from expressway maintenance road sections in Heilongjiang Province between 2002 and 2006 found that the average frequency of accidents in these sections was 2.7 times higher than that in normal sections, with a maximum ratio of 7.2, and that accident severity was 1.54 times greater than during non-maintenance periods [21]. Wang Shiwei et al. [22] investigated traffic safety incidents in tunnels and found that unsafe human behaviors and large vehicles were major contributing factors. The time periods of between 2:00 and 3:00 AM and 14:00 PM were identified as high-risk intervals, and accidents were more frequent at tunnel entrances, exits, and curved sections. The main accident types included rear-end collisions, impacts with fixed objects, side collisions, and same-direction sideswipes. Wei Mo et al. [23] retrieved 400 related studies from the CNKI database published between 1985 and 2023 on the influencing factors of road traffic accidents in China. Their analysis indicated that expressways constitute a key focus area in this research field. Similarly, Wang Fangfei et al. [24] examined accident characteristics in mountainous expressway tunnels and found that emergency parking zones and adjacent sections exhibited higher accident rates, posing serious threats to tunnel safety. Factors such as non-holiday periods, rainy weather, entrance weaving areas, large vehicles, and curved segments were identified as major contributors to severe accidents. Hu Yanting et al. [25] constructed a two-layer stacked prediction model integrating neural networks and tree models based on accident data, traffic flow data, and road segment characteristics for accident risk prediction.

Overall, the most common types of accidents in expressway maintenance road sections in China include rear-end collisions, crashes, and sideswipes between vehicles traveling in the same direction; accidents caused by improper turning or lane deviation; collisions with construction signs or isolation facilities; and impacts involving vehicles, construction machinery, or workers within the work zones. These accident characteristics are largely consistent with those observed in international studies.

3.1.2. Non-Accident Data Analysis

In recent years, some scholars have sought to move beyond the traditional safety analysis approach based solely on traffic accident statistics, adopting instead non-accident data-based safety assessment methods, among which the traffic conflict technique (TCT) has attracted considerable attention. By identifying and quantitatively analyzing potential risk events, this method enables the prediction and evaluation of traffic safety levels before actual accidents occur, thereby overcoming limitations such as the scarcity and lag of accident data. A schematic representation of the typical relationship between expressway traffic flow and a maintenance road section is shown in Figure 5.

Tian Jinyue et al. [26], based on the fluid continuum theory, utilized pipeline rheological parameters to analyze the traffic conflict characteristics in expressway maintenance road sections and established a mathematical model describing the interaction between construction machinery and high-speed traffic flow, providing a theoretical foundation for understanding traffic disturbance mechanisms in work zones. Zheng Lai et al. [27] selected Post-Encroachment Time (PET) as the conflict evaluation indicator, constructed a PET threshold-exceedance model for lane-changing maneuvers on expressways, and verified the feasibility and effectiveness of using traffic conflict techniques as an alternative to traditional accident-based methods. Meng Xianghai et al. [28] collected traffic parameters—such as vehicle type, speed, and headway distance—in two typical types of maintenance road section (half-width closures and one-way overtravel closures). Based on Time to Collision (TTC), they proposed an index for tailgating conflict frequency, and combined it with Deceleration Rate to Avoid a Crash (DRAC) to establish a quantitative model for tailgating risk assessment, enabling the evaluation of following-behavior safety in maintenance road sections. Jiang Ruoxi et al. [29] applied video-based detection and conflict recognition techniques to extract continuous microscopic vehicle motion data, constructed a traffic conflict consequence severity model grounded in vehicle collision dynamics, and used the conflict severity rate index to evaluate the overall safety level of maintenance road sections.

Overall, traffic conflict-based safety analysis provides a new perspective and technical pathway for studying traffic safety in expressway maintenance road sections. It enables a paradigm shift from “post-accident analysis” to “pre-incident risk prediction”, playing a significant role in enhancing proactive safety management under complex maintenance conditions.

3.1.3. Safety Risk Assessment

To minimize the adverse effects of potential safety hazards on expressway maintenance road sections, conducting systematic and quantitative safety risk assessments is of great significance. Yang Wen’an et al. [30] identified 36 major safety risk factors for expressway maintenance road sections through a literature review and expert consultation and further screened 17 key risk indicators. Based on the analysis results of the Structural Equation Model (SEM), these risk factors were classified into four categories—operator risk, external environment risk, management risk, and mechanical equipment risk—and their corresponding weights and priorities were determined, providing a scientific foundation for subsequent safety management. Qiang et al. [31,32] developed a Probabilistic Quantitative Risk Assessment (QRA) model based on accident frequency and consequence severity to evaluate casualty risks in long-term maintenance road sections. Incorporating traffic operation data, they introduced the Deceleration Rate to Avoid Collision (DRAC) indicator to quantitatively analyze the risk of rear-end collisions. The results showed that the risk of rear-end accidents increased significantly with the proportion of heavy vehicles and traffic volume; the lane adjacent to the work zone exhibited the highest risk, and trucks were more prone to rear-end collisions than passenger cars, indicating the significant influence of vehicle type and lane position on accident risk. In addition, Elghamrawy [33] found that the layout configuration of work zones had a substantial impact on both accident risk and associated economic losses, and that an optimized layout could effectively reduce risk exposure and accident costs. Shauna et al. [34] applied a logistic regression model to examine how different driving behaviors under merging conditions affect traffic operations and safety in work zones. The results revealed that lane-changing, deceleration, and driver attention allocation play crucial roles in safety risks, leading to the proposal of targeted safety control measures.

Wu Biao [35] adopted a structured modeling approach to identify key factors influencing traffic risk in expressway maintenance road section. In this study, the author utilized microscopic traffic flow simulation software to extract traffic operation parameters under typical maintenance scenarios, established a multivariate linear regression model to quantitatively evaluate the traffic risk of maintenance road section, and classified risk levels accordingly, thereby realizing the visualization and hierarchical management of risk levels. Zhang Xu [36] collected traffic flow data from maintenance road sections through on-site observations, analyzed the characteristics and spatiotemporal distribution of traffic operations, and proposed a quantitative safety risk analysis method based on empirical data, providing a more field-adaptive technical approach for traffic safety management. Wang Xiong et al. [37], starting from the mechanism of driving risk formation, investigated the effects of transition-zone length, traffic volume, and driving speed on driving risk. The conflict rate was selected as the core evaluation index, and the fuzzy comprehensive evaluation theory was applied for multi-factor comprehensive analysis. The results indicated that transition-zone length and traffic volume were the most significant factors influencing driving risk in the upstream transition area of expressway maintenance road sections, followed by driving speed, revealing the sensitivity of driving behavior risk to traffic environmental parameters. Zhang Yunyu [38] combined questionnaire surveys and expert consultations to identify risk indicators that significantly affect maintenance construction safety and established a systematic risk assessment index. On this basis, the Analytic Hierarchy Process–Fuzzy Comprehensive Evaluation (AHP-FCE) method was employed to quantitatively assess the safety risks of maintenance road sections, determine the weights and risk levels of various indicators, and clarify the contribution of different factors to the overall risk level.

Li Ying [39] conducted a systematic analysis of the safety issues associated with expressway maintenance road sections in Chongqing. Based on a comprehensive safety risk index system, a maintenance operation risk evaluation set was constructed, and the probability of safety risk occurrence was used to calculate risk scores for different types of operations. This approach enabled the quantitative assessment of risk levels across various maintenance activities and provided a reference for regional risk classification and management.

Liu Benmin et al. [40] collected drivers’ physiological and behavioral data—such as heart rate, speed variation coefficient, and speeding ratio—through driving simulation experiments to investigate the mechanism of driving risk formation under maintenance scenarios. Using the cumulative distribution function method, the study proposed a driving risk control evaluation model for expressway maintenance road sections and defined the safety boundary of the risk control efficiency index (C). When C < 0.3 C < 0.3 C < 0.3, the control efficiency of driving risk was classified as moderate to high, indicating that the risk was in a controllable state. This provided a scientific criterion for quantifying driving safety management.

Li Yun [41] conducted both macroscopic and microscopic analyses of safety risk characteristics in maintenance road sections. The macroscopic results showed significant variations in vehicle speeds among different regional road sections and substantial speed disparities among vehicle types, both of which increased overall traffic safety risks. The warning area, upstream transition area, and work area were identified as high-risk zones for accidents. The microscopic analysis further revealed the risk characteristics of typical traffic conflict scenarios, such as tailgating, forced merging, and compressed lane-changing, providing insight into the mechanisms of traffic conflicts under complex conditions.

Huang Fubin [42] performed a sensitivity analysis of multidimensional factors influencing driving risk in expressway maintenance road section, including traffic flow parameters and driving behavior. Using multiple linear regression, a regional risk evaluation model was developed. On this basis, the Peak-Over-Threshold (POT) model and Generalized Pareto Distribution (GPD) parameters were introduced to construct a non-accident-based quantitative driving risk evaluation model. By defining an appropriate risk threshold, the accuracy of driving risk measurement and the applicability of the model were effectively improved.

In summary, Western scholars initiated systematic studies on the safety characteristics of expressway maintenance road section relatively early, establishing a comprehensive framework integrating theoretical research, analytical methods, and practical applications. Relying on well-developed accident statistical systems, they conducted in-depth studies based on large-scale datasets. Meanwhile, Chinese researchers have actively explored safety analysis methods tailored to domestic traffic conditions, gradually shifting from traditional post-accident analysis to proactive safety assessment approaches based on traffic conflict techniques and non-accident data. Through risk identification, quantitative analysis, and model development, significant progress has been made in recognizing key safety risk factors and formulating targeted prevention strategies, providing solid technical support and scientific foundations for the safety management of expressway maintenance road sections.

3.2. Traffic Capacity of Maintenance Road Sections

Studies on the traffic capacity of expressway maintenance road sections in the United States date back to the 1970s. Early studies generally identified several key factors influencing the capacity of maintenance road sections, including work-zone length, lane-closure configuration, lateral clearance from the closed lane, proportion of heavy vehicles, gradient, construction time, and work intensity [43,44]. Building upon these foundations, numerous scholars subsequently proposed various capacity analysis and reduction models to accommodate diverse work-zone layouts and traffic conditions.

For example, Sarasua et al. [45] developed a capacity reduction model for lane-occupancy work zones, considering factors such as the location, length, and operation type of the work zone, thereby providing a quantitative basis for traffic management under different construction layouts. Yeom et al. [46] established a capacity calculation model incorporating parameters such as the number of closed lanes, lateral clearance, construction duration, and work-zone type, which can be used to predict traffic performance under various operational scenarios. Mahmoud et al. [47] designed two representative work-zone layout schemes and conducted a driving simulation experiment combined with a full-factorial analysis to systematically compare safety and capacity differences under varying traffic densities and traffic sign configurations. Nina et al. [48] statistically analyzed the impacts of the number of lanes, lane widths, proportion of heavy vehicles, lane reduction conditions, percentage of commuter vehicles, and lane splitting on work-zone capacity. Using the speed–flow model and stochastic capacity theory, they calculated the long-term traffic capacity of maintenance road sections and estimated short-term capacity through the queuing flow rate after congestion. Lu et al. [49] further integrated the speed–flow relationship and the characteristics of free-flow speed to construct a predictive model for work-zone capacity, which was also applied to analyze the formation and duration of traffic congestion.

Overall, Western scholars have developed a systematic theoretical foundation and computational framework for studying expressway maintenance work-zone capacity. Their research methods encompass field data analysis, driving simulation, statistical regression, and stochastic modeling, providing a robust foundation for optimizing traffic organization and enhancing construction safety management.

In China, researchers have primarily relied on microscopic traffic simulation techniques to investigate the traffic capacity of expressway maintenance road sections and the factors influencing it. According to existing studies [50,51,52], the traffic capacity of maintenance road sections is jointly affected by multiple factors, including the passage organization scheme (or lane-closure configuration), lane width, proportion of heavy vehicles, speed limit, longitudinal gradient, and the length of the work zone.

Li Jie [51], using microscopic simulation, systematically analyzed the influence of various factors on the traffic capacity of maintenance road sections by applying reduction rates under a unified reference index. The results indicated the following order of importance: passage organization scheme > proportion of heavy vehicles > work-zone length > speed limit > longitudinal gradient. Feng Minxia [52] conducted similar simulation-based research and obtained consistent findings, ranking the influencing factors as follows: lane-closure type > work-zone length > proportion of heavy vehicles > lane width > speed limit > longitudinal gradient. Both studies demonstrated a high degree of consensus among scholars regarding the primary determinants and relative significance of influencing factors—namely, that the traffic organization pattern and the proportion of heavy vehicles are the key variables governing work-zone capacity.

Liu Lei [53], based on the Cellular Automata (CA) theory, developed traffic flow and lane-changing models for expressway maintenance road sections and independently programmed a simulation tool on the MATLAB R2017a platform. By varying parameters such as the proportion of heavy vehicles, work-zone length, and lane-closure configuration, they conducted multiple comparative simulations to reveal the dynamic patterns of traffic operations and capacity variations under different conditions, providing a reference for the application of simulation modeling in work-zone traffic analysis.

Additionally, Pei Yulong et al. [54] integrated the concept of Intelligent Transportation Systems (ITSs) to design an intelligent lane-merging control system for expressway maintenance road sections. The system can dynamically select optimal merging strategies based on real-time congestion levels, thereby improving vehicle throughput while reducing conflict risks and safety hazards within the work zone—achieving coordinated optimization of both safety and capacity.

In summary, the traffic capacity of expressway maintenance road sections is influenced by a combination of factors, including roadway conditions, construction organization, traffic composition, speed and control measures, and environmental factors. Through a combination of theoretical analysis, field observation, and traffic simulation, both Chinese and Western scholars have systematically explored the underlying mechanisms and influencing patterns. These studies have not only deepened the understanding of traffic operation characteristics in maintenance road sections but also provided a solid theoretical and technical foundation for developing effective traffic control strategies, delay evaluations, and capacity optimization measures.

3.3. Work-Zone Layout of Maintenance Road Sections

In China, maintenance road sections are primarily classified into seven components: the advance warning area (S), the upstream transition area (Ls), the longitudinal buffer area (H), the transverse buffer area (Hh), the activity area (G), the downstream transition area (Lx), and the termination area (Z), as shown in Figure 6.

Both Chinese and Western authorities have developed a series of standards and technical specifications governing the layout of expressway maintenance road sections. For instance, in the United States, the Manual on Uniform Traffic Control Devices (MUTCD) and the Best Practices Manual for Work Zones provide systematic guidelines for traffic signage placement, lane-closure configurations, and safety control measures within work zones. In China, the primary standards include the Safety Work Rules for Highway Maintenance (JTG H30-2015) and Road Traffic Signs and Markings—Part 4: Work Zone (GB 5768.4-2017), which specify requirements for zoning layout, safety facility deployment, and traffic organization in maintenance operations.

However, in order to ensure generality and operational feasibility, existing standards primarily focus on macro-level principles for work-zone layout design, while paying comparatively limited attention to detailed influencing factors such as traffic capacity, accident frequency, work scale, and environmental variability. As a result, in practical applications, some work-zone layout schemes fail to fully align with specific traffic conditions and construction constraints, often leading to reduced service levels, aggravated traffic bottlenecks, and increased driving safety risks, thereby undermining the overall operational efficiency and socio-economic benefits of expressways.

3.3.1. Optimization of Work-Zone Length

According to the maintenance requirements and operational scope of expressway projects, different types of maintenance sections generally correspond to varying design length requirements. The design length of each functional area—such as the warning area, upstream transition area, and work area—directly affects vehicle passage efficiency and driving safety, and the rational configuration of these three areas is particularly crucial.

Internationally, scholars have widely adopted an approach that integrates comprehensive cost analysis with simulation modeling to systematically optimize the length of work zones. McCoy et al. [55] developed an optimal work-zone length model for lane closures on conventional four-lane expressways, aiming to minimize the adverse impact of maintenance activities on expressway operational efficiency and overall operating costs while meeting construction requirements. Similarly, Steven et al. [56] studied the optimal length configuration of four-lane expressway work zones based on the minimum total cost principle, verifying the effectiveness of cost-constrained length optimization strategies. Ceder [57] established a comprehensive cost optimization model for two-lane expressways, balancing traffic delay costs with maintenance operation costs to determine the optimal work-zone length under different traffic flow conditions.

In addition, Lee et al. [58] introduced the Traffic Conflict Index (TCI) to evaluate the optimal length of transition areas in urban road maintenance sections, proposing a safety-oriented optimization approach that emphasizes the leading role of safety performance indicators in work-zone design. Yeom et al. [59] constructed a traffic simulation model for expressway maintenance road sections to analyze the configuration of work-zone lengths under various lane arrangements and closure conditions, providing methodological support for simulation-based optimization of work-zone design.

Overall, Western research on work-zone length optimization emphasizes the coordinated balance between traffic efficiency, construction safety, and economic cost. Most studies employ a combined cost modeling and simulation analysis framework, enabling quantitative and dynamic optimization of work-zone length. These findings provide important theoretical foundations and methodological guidance for the scientific design and management of expressway maintenance road sections.

In China, Li Xiaolong [60] conducted a statistical analysis of traffic accident data from expressway maintenance road sections. Considering multiple factors such as construction intervals and nighttime operations, a total maintenance cost function model was established, and the optimal open length of work zones under different construction organization conditions was investigated, providing theoretical support for balancing construction efficiency and traffic safety. Yi Yanfeng [61] systematically studied the length calculation methods and design standards for different functional zones—such as the warning area, transition area, buffer area, and termination area—and proposed a quantitative model for determining reasonable work-zone lengths based on cost–expense functions and simulation techniques, achieving greater precision and operability in design.

Peng Yuhua et al. [62] used VISSIM simulation software to model traffic conditions during expressway maintenance, employing saturation, traffic density, average speed, and average delay as comprehensive evaluation indicators to optimize the service level of work zones. A dynamic comprehensive evaluation method was further applied to reveal the variation patterns of work-zone length under different service levels, providing data-driven support for dynamic traffic management. Wei Lulu [63], from the perspective of driver behavior and cognitive demand, analyzed the reasonable length settings of temporary maintenance road sections and proposed a coordinated design approach integrating work-zone length optimization with the deployment of safety facilities, thereby improving both driving safety and traffic flow stability. Li Daofu [64] constructed a game-theoretical model describing the interactions between lane-changing and target vehicles within maintenance road sections and examined the optimal length design of warning areas under different design speeds, achieving coordinated optimization between driving behavior and traffic safety.

Wu Jiangling et al. [65] found that the layout configuration of work zones significantly affects the cost of traffic accidents. Based on the principle of accident cost minimization, an optimization model for work-zone layout was developed, and case analysis was conducted to identify characteristic parameters associated with the highest risk levels. Wang Shouwei [66], focusing on eight-lane expressways, proposed 27 layout configurations corresponding to different closed-lane scenarios. Using Principal Component Analysis (PCA) for comprehensive simulation evaluation and combining rapid clustering and hierarchical cluster analysis, the study revealed that the warning area and upstream transition area are the critical regions influencing work-zone traffic capacity, determining overall operational efficiency and safety performance.

In summary, Chinese scholars have established a relatively systematic technical framework for work-zone length optimization and layout design, encompassing accident data analysis, cost modeling, and traffic simulation validation. The research findings demonstrate that scientifically optimized work-zone length design can not only improve traffic efficiency and service levels but also significantly reduce driving risks, providing a solid theoretical foundation and practical guidance for the refined design, dynamic optimization, and intelligent management of expressway maintenance road sections.

3.3.2. Optimization of Safety Facilities

The installation of diverse safety facilities in expressway maintenance road sections serves a dual purpose. Firstly, it aims to ensure the safety of construction personnel and equipment. Secondly, it functions to offer reminders, warnings, isolation, and guidance, thereby guaranteeing the safe passage of vehicles. Figure 7 depicts a layout schematic of safety facilities in each area of the expressway maintenance road section.

The scientific deployment of safety protection facilities is a vital measure for reducing potential risks and safety hazards, lowering accident frequency, and improving both traffic safety and operational efficiency in expressway maintenance road sections. The Manual on Uniform Traffic Control Devices (MUTCD) in the United States provides systematic principles and technical standards for the design, application, and deployment of safety facilities within the maintenance road section during maintenance work [67].

Azuma et al. [68], through field investigations and data analysis, identified that among the key factors affecting work-zone safety, the comprehensibility of traffic signs is the most critical, followed by nighttime visibility. Their study found that the readability and recognizability of signs can be significantly improved by using pictorial symbols and avoiding boldface fonts, while floodlighting can enhance the nighttime visibility of traffic controllers and safety devices, effectively reducing potential traffic conflicts.

Rahman et al. [69] focused on the design and application of Dynamic Message Signs (DMSs), examining the effects of DMS content, frame refresh rate, and installation position on driver deceleration behavior. The findings revealed that while content and position had limited impact, the frame refresh rate of DMSs had a significant effect on driver deceleration response. Furthermore, DMSs were found to be more effective at night than during daytime, underscoring the importance of dynamic information updates under low-visibility conditions.

Bahram et al. [70], from the perspective of police deployment, analyzed how different levels of law enforcement presence influence traffic safety in work zones. By using indicators such as speed variation, average speed reduction, proportion of high-speed vehicles, and the 85th percentile speed, they quantitatively assessed the safety control effectiveness of enforcement. The results demonstrated that appropriate police deployment effectively reduces speeding rates and traffic conflict occurrences, significantly enhancing the overall safety performance of maintenance road sections.

Overall, Western scholars have developed a relatively comprehensive work-zone safety facility optimization framework, emphasizing perspectives of traffic sign visibility, information transmission efficiency, and external safety interventions. This framework highlights the transformation from “passive protection” to “proactive prevention”, integrating multi-dimensional perception, real-time feedback, and behavioral guidance mechanisms to substantially improve the proactive safety capacity and overall safety level of expressway maintenance road sections.

In China, expressway maintenance road sections are typically equipped with a wide range of safety protection facilities, including warning signs and markings, traffic cones, crash barrels, warning lights, variable message boards, crash cushions, and night lighting devices. However, in practice, a considerable number of accidents are still closely associated with improper or insufficient deployment of safety facilities, indicating that the current protection system requires further improvement in terms of precision design and dynamic responsiveness.

Chen Yu [71] investigated the visual recognition characteristics of drivers toward traffic signs in expressway maintenance road sections, proposed methods for setting various types of signs, and developed a systematic design framework for traffic organization schemes, providing methodological support for scientific layout and guidance strategies. Yu Ying et al. [72] analyzed traffic conflicts and operational characteristics, established a probabilistic safety prediction model for maintenance road sections, and derived quantitative relationships for the deployment and utilization of safety facilities, laying a theoretical foundation for systematic safety management.

Yi Yanfeng [59] conducted an integrated study of traffic signs, markings, lane separation facilities, crash barriers, and speed control systems, establishing a visual recognition model for traffic signs and proposing graded speed limit standards and corresponding control measures. Using the Analytic Hierarchy Process (AHP) and Fuzzy Comprehensive Evaluation (FCE) methods, the study quantitatively evaluated the protective effectiveness of safety facilities, verifying the synergistic effects among multiple safety measures.

Zhou Jin [73], focusing on drivers’ visual cognition and eye movement patterns, analyzed attention allocation mechanisms during sign recognition and validated the rationality and effectiveness of signage design and layout through driving simulation experiments. Zhu Fuchun [74] innovatively proposed a three-dimensional color optical illusion deceleration marking, suggesting its combined use with vibration deceleration markings to enhance drivers’ perception of speed and deceleration cues, thereby improving safety warning performance in work zones.

Cheng Pengfei [75] provided practical recommendations for the deployment of guidance and protection facilities in expressway maintenance road sections and determined the optimal spacing of traffic cones through experimental studies. Wu Wenyan [76] used driving simulation to collect vehicle operation data and drivers’ psychological responses, and then constructed an improved Metacellular Automaton traffic flow model based on driver tension levels. This model was used to determine the optimal advance lane-change position and the spacing and placement of speed limit signs under varying service levels, achieving coordinated optimization between safety and traffic efficiency.

Bochao Li [77] examined the influence of smart stud clusters on driver behavior through driving simulation experiments, finding that the warning distance of the clusters significantly affects drivers’ lane-change timing and local speed control. Liu Benmin [38] designed a multi-group comparative experiment involving voice prompts, different speed limit levels, and ground markings, and combined driving simulation with questionnaire surveys to evaluate the comprehensive effectiveness of multi-modal control measures, confirming that a multi-modal early warning system can substantially enhance safety performance in maintenance road sections.

In conclusion, despite the late start of China’s research on the optimization of the work-zone layout of expressway maintenance road sections, it has made significant progress in recent years—particularly in the length optimization of functional areas and the optimization of safety facility configurations—and has gradually formed a more systematic research framework. Unlike Western scholars who rely on long-term, large-scale datasets from maintenance road sections for systematic modeling and statistical analysis, Chinese researchers have focused on engineering-oriented and problem-driven approaches, using traffic simulation, field observation, and experimental validation to explore optimization strategies suited to the nation’s specific traffic and construction conditions.

Current studies primarily concentrate on specific regional contexts or representative scenarios of expressway maintenance road sections, as well as the layout optimization of key functional areas such as the advance warning area, the upstream transition area, and the work area. These studies not only reveal the mechanisms of traffic flow evolution and risk variation in complex construction environments but also provide scientific and practical foundations for improving safety performance, traffic capacity, and organizational efficiency in maintenance zones. Overall, China’s research in this field is evolving from empirical exploration toward data-driven modeling and intelligent decision-making, laying a solid foundation for the refined design, intelligent management, and high-efficiency operation of expressway maintenance road sections.

3.4. Speed Limit Management of Maintenance Road Sections

According to related surveys, in accidents occurring on expressway maintenance road sections, forced overtaking, speeding, and improper driving operations account for 5.95%, 29.88%, and 10.45%, respectively—most of which stem from unreasonable speed-reduction settings [17,78]. It is worth noting that lowering the posted speed limit does not necessarily reduce actual vehicle speeds, nor does it always decrease speed variance, indicating that the scientific rationality and dynamic adaptability of speed limit management are decisive for maintaining traffic safety.

Generally, speed limit management in expressway maintenance road sections can be divided into two main categories: static speed limit management and dynamic (variable) speed limit management. The former relies on fixed speed limit signs to prompt drivers to slow down—simple in structure but lacking responsiveness to real-time traffic conditions. The latter, in contrast, integrates real-time traffic parameters, weather conditions, and construction activity information to dynamically adjust speed limits, offering superior adaptability to complex and changing environments and achieving a better balance between safety and efficiency.

3.4.1. Static Speed Limit Management

Most studies on static speed limit management for expressway maintenance road sections primarily focus on two core issues: the reasonable determination of speed limit values and the optimization of speed limit sign placement. Early research found that the average speed of vehicles passing through maintenance road sections decreases significantly with an increase in construction intensity, and the closer the work zone is to the traffic lane, the lower the average vehicle speed. When traffic flow is relatively low, this effect becomes less pronounced. Moreover, extensive field data indicate that the actual vehicle speeds through work zones are often higher than the posted speed limits, and this discrepancy is closely related to the number of traffic lanes [79].

Richards et al. [80] pointed out that speed control is a key measure for enhancing traffic safety in maintenance road sections. Properly set speed restrictions can effectively reduce the dispersion of traffic speeds, promote uniformity in flow, and thus lower the risk of accidents. James et al. [81] established a quantitative relationship model between speed limits and safety levels from a traffic safety perspective, providing a theoretical basis for the scientific formulation of speed limit standards. Saito et al. [82] empirically evaluated the effectiveness of speed management measures and found that the addition of speed-measuring devices and police enforcement could reduce the average vehicle speed by approximately 6 mph, thereby significantly improving the safety level within the work zone.

Wu et al. [83] analyzed the impact of the proportion of heavy vehicles on traffic flow speed and speed dispersion within work zones based on the dynamic characteristics of different vehicle types. Results showed that an increase in heavy vehicle proportion exacerbates speed differences, thereby increasing the likelihood of traffic conflicts and accidents. Dai Tongyu et al. [84], based on principles of human factors engineering, quantified drivers’ information-processing behavior and proposed a speed limit calculation model considering driver cognitive load, achieving a balance between speed limit design and driver perceptual capacity. Feng Zhicheng [85], grounded in the principle of energy conservation, established a speed limit regulation model for typical maintenance road sections, optimized the placement of speed limit signs, and identified the optimal control strategy for advance warning zones through simulation analysis.

Rao Xiangru [86], taking the closure of the outer two lanes of a six-lane expressway as the study object, proposed three typical speed limit modes, F-S (fast–slow deceleration), M-M (uniform deceleration), and S-F (slow–fast deceleration), and compared them using safety and service level as evaluation indicators. Xia Ran [87] proposed an improved traffic conflict identification and prediction model that can effectively detect and predict rear-end collision risks in maintenance road sections and, based on this, developed a tiered speed limit control strategy that significantly enhances traffic safety.

3.4.2. Variable Speed Limit Management

The application of Variable Speed Limit (VSL) technology in expressway maintenance road sections is increasingly prevalent. Current research primarily focuses on the determination of VSL thresholds, design of control algorithms, evaluation of effectiveness, and optimization of sign placement, with the overall goal of achieving adaptive speed management strategies that respond dynamically to complex and changing traffic conditions.

Lee et al. [88] investigated the effects of different VSL strategies on traffic risk and proposed a real-time VSL theoretical model capable of reducing accident rates by approximately 5–17%. Their findings indicated that the optimal safety benefit is achieved when the VSL value aligns with the average vehicle speed in the transition area. Papageorgiou et al. [89], using a dynamic traffic flow model, systematically analyzed the characteristics and operational states of expressway traffic flow and verified the effectiveness of VSL strategies in smoothing traffic flow, reducing speed fluctuations, and mitigating risk. Joy et al. [90] proposed a feedback control-based VSL algorithm, which was validated through simulation to improve both safety and traffic efficiency under multi-stream traffic conditions. Huang et al. [91] conducted field experiments in the upstream areas of expressway maintenance road sections, revealing that Portable Changeable Message Signs (PCMSs) can reduce average vehicle speeds by approximately 13–17%, thereby significantly improving safety performance within the work zone.

In China, Ma Qiannan [92] developed optimal VSL control models corresponding to high-density and low-density traffic flow conditions, achieving adaptive optimization of speed limit strategies under varying traffic states. Liu Borong [93] compared VSL signs with fixed speed limit signs in the advance warning area of expressway maintenance road sections and found that the VSL signs demonstrated superior control effectiveness and safety enhancement. Furthermore, an information-benefit maximization model was employed to optimize the spatial configuration of sub-control zones, improving the precision and responsiveness of the control strategy. Liu Xu [94] conducted microscopic traffic simulations to compare dynamic and static speed limit control, revealing that vehicle-type-specific VSL control can significantly enhance traffic flow stability while maintaining both safety and efficiency.

In summary, both static and dynamic speed limits serve as essential management measures for expressway maintenance road sections. While static speed limits stabilize vehicle speeds and reduce driver violations, dynamic VSL systems integrate real-time traffic data, weather conditions, and construction activity to flexibly adjust control strategies, thus ensuring safety while further improving traffic efficiency. Empirical evidence suggests that under complex and fluctuating traffic environments, dynamic VSL systems often outperform static ones in achieving comprehensive performance. Future developments should therefore emphasize data-driven, regionally coordinated, and precision-based speed management, enabling the adaptive, intelligent, and safe operation of expressway maintenance road sections.

4. Conclusions

The issue of traffic safety in expressway maintenance road sections has drawn widespread attention from scholars and transportation authorities both in China and in Western countries. Existing studies indicate that this field encompasses a broad range of topics—including safety risk identification and assessment, analysis of traffic capacity, optimization of work-zone layout, and speed limit management—and is not limited to the four dimensions summarized in this paper. Overall, Western researchers have established more systematic and mature frameworks for both theoretical and practical applications. In contrast, Chinese studies primarily rely on field investigations and traffic simulation methods, focusing largely on specific maintenance scenarios; thus, their results often exhibit limited generalizability and applicability.

Because of substantial differences between China and Western countries in expressway construction planning, technical standards, development philosophies, traffic composition, and construction scale, the safety and management characteristics of maintenance road sections differ markedly. Consequently, while Western research provides valuable reference material, its findings cannot be directly transplanted to Chinese conditions. In summary, although China has achieved notable progress and phased results in this field, further in-depth and innovative exploration remains necessary in several key directions.

(1) Compared with the long-term accumulation and well-established databases on maintenance safety and accident statistics abroad, China still exhibits relative deficiencies in systematic and standardized data collection. Consequently, most existing research findings are confined to localized projects or specific sections, lacking broad applicability. With the rapid advancements in smart expressways and intelligent maintenance technologies, it is essential to establish a traffic safety data collection and management platform tailored to maintenance operation scenarios. This platform should enable the integrated acquisition, standardized management, and multidimensional analysis of accident, operational, and environmental data. By incorporating data-mining techniques and artificial intelligence-based modeling, a solid data foundation can be provided for research on the safety mechanisms and systematic traffic safety management of expressway maintenance road sections.

(2) With the continuous increase in traffic volume, the number of expressway with eight or more lanes has grown significantly, and in some regions, super-scale expressways with ten to twelve lanes have already been constructed. These sections are subject to multiple influencing factors during routine maintenance and medium-to large-scale rehabilitation operations, resulting in a highly complex traffic environment and elevated risk exposure. Moreover, the existing layout models in current standards can no longer fully meet the practical requirements of such large-scale expressways. Once an accident occurs, its impact range and consequences tend to be far more severe. Therefore, future research should systematically investigate the traffic characteristics, risk mechanisms, traffic capacity, and speed limit control strategies of maintenance road sections on super multi-lane expressways, and further improve the corresponding safety management standards and design guidelines.

(3) Existing studies primarily focus on ordinary straight segments of expressway maintenance road sections, while special sections—such as interchange ramps, bridges and tunnels, long steep gradients, mountainous roads, and areas with high geological hazard risks—have received insufficient attention. Furthermore, research on traffic safety under extreme climatic and special operating conditions (e.g., high-temperature exposure, snow, rain, fog, strong winds, dust storms, nighttime, or low-illumination environments) remains relatively limited. Future studies should integrate field data, simulation technologies, and intelligent perception methods to systematically elucidate the traffic operation characteristics and accident risk evolution mechanisms under complex environmental conditions, thereby providing a theoretical foundation for safety protection and management of maintenance road sections under special conditions.

(4) In recent years, studies based on driving simulation experiments and behavioral cognition models have been increasing; however, most of them focus on driver state recognition and road environment perception, while systematic research on maintenance road sections remains limited. From the perspective of the traffic system, the interaction between drivers’ subjective cognition and external environmental factors constitutes a key mechanism influencing safety in maintenance areas.

Future research should integrate next-generation driving simulation platforms, big data analytics, and artificial intelligence technologies to develop intelligent simulation models that incorporate the stochasticity, complexity, and temporal variability of driving behavior. Such models should aim to uncover the spatiotemporal evolution patterns of traffic conflicts and risks under various layout schemes, thereby providing a theoretical foundation for intelligent risk assessment and proactive safety control in expressway maintenance road sections.

(5) With the increasing number of early-built expressways entering the expansion and reconstruction stage, issues such as traffic congestion, speed reduction, and frequent accidents have become increasingly prominent. Compared with conventional maintenance operations, expansion projects are characterized by larger scale, longer construction periods, and more complex traffic organization forms, involving the multi-dimensional optimization and coordination of work-zone layout, functional area length design, safety facility configuration, and speed limit management.

Although some domestic studies have preliminarily explored the traffic safety issues associated with expressway expansion sections, a comprehensive and systematic standard and technical framework has yet to be established. Therefore, future efforts should strengthen the integration of theoretical research and engineering practice, and establish a standardized system for traffic safety evaluation and design during expressway expansion and reconstruction. This will provide essential support for ensuring safe operation and effective management under complex and dynamic construction conditions.

Overall, as China’s expressway network enters a stage of comprehensive maintenance and intelligent transformation, intelligent, digital, and refined maintenance management will become the prevailing direction of industry development. In the future, traffic safety issues under intelligent maintenance scenarios should be incorporated as a key research focus. Relying on information perception, vehicle–infrastructure cooperation, and intelligent decision-making technologies, it will be possible to achieve dynamic optimization and adaptive regulation of safety facilities and traffic management. This will comprehensively enhance the safety protection capacity, operational efficiency, and service quality of expressway maintenance road sections.