1 Introduction
Cities are places of diversity and complex interactions (
Harvey, 2012) that if not planned and managed properly can become unsustainable (
Gao et al., 2023;
Tsagkis et al., 2023). While cities should provide a livable setting for people and goods, it is evident that urban ecosystems have historically been designed to prioritize vehicles over people (
Lambrianidou et al., 2013). In this context, the structure and configuration of urban and transport networks play crucial roles in shaping mobility behavior. It is important to note that these networks are not homogeneous; rather, they exhibit hierarchical order (
Lammer, Gehlsen, Helbing, 2006;
Shang et al., 2020). According to Batty (
2006), hierarchy is inherent in city systems, regardless of whether they are organic or ‘microwaved’. Hierarchy is a multidimensional concept that can be interpreted in four ways (
Lane, 2006): a)
order hierarchy, which depicts an “ordering induced by the values of a variable defined on some set of elements”; b)
inclusion hierarchy, which refers to a recursive organization of entities; c) control hierarchy, which is related to (social) power such as an army or a political party; and d)
level hierarchy, which describes “a particular kind of ontological organization, in which entities are posited to exist at different levels.”
When looking into the transport dimension, it is worth emphasizing that the majority of roads within a network are considered trivial, whereas a minority are vital (
Jiang, 2009). This distinction implies the existence of two types of hierarchy: order hierarchy and level hierarchy. Road network hierarchy refers to the classification of roads and streets into groups or classes based on specific characteristics such as form, use/function, relation, and designation (
Marshall, 2005). This definition embodies order hierarchy and can also be interpreted as street network classification (
Ribeiro, 2012). Similarly, Eppell, McClurg, Bunker (
2001) described road hierarchy as a means of assigning each roadway a function to establish appropriate objectives and design criteria. Another definition, related to level hierarchy, suggests that road network hierarchy involves assigning a role or level to each road or street segment within the entire network (
FHWA, 2013). These definitions collectively illustrate that road network hierarchy involves the process of ordering, classifying, and categorizing road networks.
The road network hierarchy can be either conventional (car-oriented) or alternative (multimodal and sustainable modes oriented). This difference leads to various road environments and urban landscapes (
Marshall, 2005). The first classification plan was developed by Buchanan in the UK in 1963; he proposed two contrasting functions for the road network: movement/circulation and access (
MoT, 1963). This approach resulted in three primary road categories:
arterials, collectors or distributors,
and local roads. Arterials emphasize car movement, while local roads prioritize property access (
Levinson and Krizek, 2008). However, Buchanan’s once progressive idea has become an outdated conventional approach, which has resulted in dysfunctional urban areas with significant societal and environmental issues (
Marshall, 2004). Specifically, this car-oriented approach overlooks the social aspect of streets (
McCann, 2013), creates barriers within the urban fabric, and leads to environmental degradation (
Leurent and Windisch, 2011). In other words, it undermines the role of other modes of transport (
Liu et al., 2017). Therefore, cities should adopt innovative, efficient, and inclusive strategies to enhance urban environmental conditions (
Aletà et al., 2017;
Tan et al., 2023). Establishing a culture of sustainable mobility in cities is a challenging task that requires a paradigm shift (
Sdoukopoulos et al., 2019;
Tzouras et al., 2023) built on an integrated and multimodal urban and transport planning approach (
Bigotte et al., 2010;
Melkonyan et al., 2020). New classification schemes should focus on promoting sustainable mobility, embracing multimodality, and reducing the dominance of cars (
Stamatiadis et al., 2023;
Tsigdinos et al., 2023). By prioritizing alternative modes of transport, these classification schemes seek to create new urban and road environments that are fair and accessible to all people (
Marleau Donais et al., 2019).
Therefore, these schemes can be used in the future. Revising the urban network hierarchy will shape the future of transport (
Tsigdinos et al., 2022). New modes, technologies, and measures should be integrated into a single sustainable hierarchy/classification system. Failure to do so will pose significant challenges to urban environments, as each new and existing mode of transport will compete for dominance. On the other hand, sustainable and inclusive transport futures will prioritize the integration of all modes of transport, promote alternative modes, and encourage multimodality and coexistence. This approach has the potential to cultivate conditions for a truly viable transport future. Therefore, the direction of road hierarchy (conventional or alternative) and its role are crucial factors that determine the future of transport in urban areas.
Rethinking the current state of transport is a complex task, particularly when considering the hierarchy of urban road networks. This raises a number of pertinent questions: a) What exactly is the urban road network hierarchy? Is it purely a planning tool or does it include a broader concept? b) Is it qualitative or quantitative? c) Does it emerge naturally and as a dynamic process or is it solely the result of proactive planning? Additionally, in addition to understanding the nature of hierarchy itself, identifying prevailing trends and concepts that inform road network hierarchy schemes is also crucial. As these research questions have not yet been comprehensively and definitively answered, this study aims to explore the literature and provide a detailed interpretation of the nature of the urban road network hierarchy, thereby identifying challenges, opportunities, and perspectives that influence and are influenced by it.
Consequently, the objective of this study is to identify research papers published in Scopus-indexed scientific journals that address various aspects of street classification (such as methods, categorization definitions, and analysis). By doing so, we aim to provide clear and concise answers to the aforementioned questions. Our work adopts a systematic literature review (SLR) approach, which enables us to trace the evolution of scientific dialog on road network hierarchies over time. Furthermore, this process allows for a deeper examination of critical issues that shape the characteristics and properties of road hierarchy in urban environments. Through this systematic process, we not only draw valuable conclusions about current approaches and theoretical perspectives regarding hierarchies but also outline directions for future research-an essential endeavor for planning and delivering sustainable urban futures.
The remainder of the paper is structured as follows: the second section outlines the research methodology, detailing the fundamental steps undertaken to conduct the literature review; the third section presents the findings of the SLR process, including the results from the bibliometric analysis; the fourth section critically discusses the findings and presents policy recommendations and suggestions for future research; and finally, the last section offers concluding remarks based on our research.
2 Research methodology
SLR, is a crucial technique in social science research. It involves systematically gathering and analyzing all available information on a specific topic or effect (
Lavissière et al., 2020). The SLR serves as a method for identifying and critically evaluating relevant research (
Kraus and Proff, 2021) as well as for collecting and analyzing data from that research (
Liberati et al., 2009). Unlike traditional literature reviews, SLRs adhere to a set of principles to minimize potential biases in the sample of studies (
Booth et al., 2016). This method is effective at synthesizing research findings and providing evidence at the meta-level (
Snyder, 2019). Therefore, this approach is suitable for identifying key concepts related to the future of urban road network hierarchy. In this study, we utilize SLR to determine a solid theoretical foundation for the various concepts that may shape road hierarchy.
2.1 SLR process
In conducting the SLR process, this study adopts a three-step procedure, consistent with the approach used by Bask and Rajahonka (
2017), Yigitcanlar et al. (
2019), and Oliveira et al. (
2017). The first step is the planning stage, which involves defining the objectives and review protocol for the systematic review and identifying the sources and procedures for literature searches. The second step refers to the conduction of the review process by applying the defined inclusion and exclusion criteria. Finally, the third step involves identifying concepts through a synthesis of the results in line with the established objectives.
2.1.1 Keyword selection
First, a research plan was developed, outlining the research aim, questions, keywords, and a set of inclusion and exclusion criteria. The research aim was framed to identify concepts and approaches related to the hierarchy of urban road networks or street classification. Accordingly, a diverse set of keywords was compiled. These keywords were used to search for relevant articles in the titles and abstracts. The keywords used included “road classification” or “street classification,” “street functional classification,” “road network hierarchy,” “street hierarchy,” “street type,” “urban road classification” or “urban roadway classification,” “street reconfiguration,” “street functionality,” “street function,” “street planning,” “road categorization,” “road categories” or “street categories,” “street typology,” “strategic road network,” “complete streets” or “road type.” All these keywords pertain to the process of classifying streets or roads.
The selection of keywords in a SLR process is of utmost importance. These keywords play a crucial role in determining the results of the method (
Liao et al., 2017b). The relevance and specificity of the keywords to the subject under consideration are directly proportional to the efficiency of the process, enabling a thorough and meaningful analysis of the issue at hand, both in the present and in the future. It is evident that the list of keywords adopted in this research aims to provide a comprehensive background. In essence, these keywords comprehensively cover the topic of urban road network hierarchy, as documented in the literature. The selection process was straightforward and transparent; initially, a concise narrative review of papers related to road network hierarchy or similar terms was conducted, incorporating some trial keywords. This review process lasted for approximately one week, resulting in the identification of a wide range of relevant terms that formed the basis of the keywords employed in this study. Hence, these keywords are deemed appropriate because they reveal the emerging trends and sub-concepts related to the urban road network hierarchy. It is worth underlining that both “road” and “street” were included as keywords to capture a diverse range of scientific publications. In fact, within the literature, these terms are often used interchangeably and synonymously (
Ershova and Smirnov, 2017). Additionally, these keywords are not accompanied by any other words, as the focus lies solely on classification and hierarchy rather than other specific characteristics or features.
2.1.2 Defining inclusion and exclusion criteria
The inclusion criteria were determined to be academic journal articles (research articles or reviews) that were available online in full text and published in English and were relevant to the research objectives, specifically the urban road network hierarchy. These articles are considered useful contributions and provide versatility, diversity, and generalizability to our study. Additionally, only papers published after 2000 were selected, in alignment with other approaches of reviewing papers that use a specified date as a threshold for efficiency and timeliness (
e.g., Tsigdinos et al., 2022). This decision allows us to capture the most recent trends and contextualize hierarchies through the lens of sustainable transport futures. Moreover, it ensures feasibility and facilitates a meaningful interpretation of the selected corpus. A preliminary examination revealed that more than 82% of the articles related to road hierarchy or classification were published after 2000, which is why the year was chosen. “Gray literature,” such as conference proceedings, books, and technical reports, was excluded from the corpus to ensure the high quality of the studies analyzed (
Lagorio et al., 2016). The search was conducted using the Scopus database.
In step 2, the search for relevant articles was carried out in April 2023, and the entire procedure lasted approximately one and a half months. The previously mentioned keywords were used in the search to identify articles that address concepts related to the transformation of the urban road of the future. Initially, the search returned a total of 2908 articles, including journal articles and (peer review) conference proceedings in all possible languages found in Scopus. After limiting the search to papers published between 2000 and 2022, the number of papers was reduced to 2392. Articles published in 2023 were not considered to ensure research consistency by only utilizing full years. Further narrowing the sample to include only journal articles (research articles or reviews) written in English resulted in a total of 1512 papers eligible for title and abstract screening.
2.1.3 Eye-balling
The selected papers were “eye-balled” through a rigorous title and abstract screening process to ensure the consistency and accuracy of the keywords used (
Bhimani et al., 2019). The papers with relevant abstracts were included in the next stage, resulting in a sample size of 147 articles. This task was challenging, so the preselected papers were divided into three groups (504 papers per group), and each author screened one group independently. This approach allowed for high flexibility and efficient handling of the preselection stage. The screening process took approximately three weeks. Notably, many of the initial articles included keywords related to road network hierarchy in a general or superficial manner rather than truly addressing the concept of hierarchy itself. This expedited the screening process. The full texts of the initially screened articles were then independently reviewed by all three authors based on the research aim. This led to a pool of 39 articles for which all three authors agreed were highly relevant. These articles were thoroughly read, categorized, and analyzed in terms of hierarchy and classification concepts. The literature selection procedure is presented through a PRISMA diagram (Fig.1), which ensures transparent and comprehensive reporting of the SLR process (
Liberati et al., 2009).
In step 3 (reporting and concept identification), the focus was on writing up and presenting the findings, specifically the critical concepts, approaches, and methods related to urban road network hierarchy. A synthesis process was conducted to highlight the fundamental elements and concepts associated with hierarchy. Additionally, relevant concepts were formulated by combining different elements found in the literature. It is worth underlining that other publications related to hierarchy and classification in urban areas were also utilized at this stage. As a result, the snowballing method, commonly used in scientific literature reviews (
Lagorio et al., 2016;
Lagorio et al., 2022;
Tsigdinos et al., 2022), increased the number of reviewed references to 42 (an addition of 3 scientific articles).
3 Results
This section presents the research results, including descriptive and detailed bibliometric analyses. It provides an overview of the sample characteristics, identifies existing trends and their future potential, explores concepts related to hierarchy, and presents a citation network analysis to illuminate the interactions between the current literature components.
3.1 Bibliometric analysis: Describing the corpus
First, when it comes to contextualizing and understanding the research agenda on a specific topic, it is essential to observe the development of the corpus of literature over the years. A bibliometric analysis allows us to unravel the evolutionary nuances of a particular field and shed light on emerging areas within that field (
Donthu et al., 2021). Therefore, a time series illustrating the number of publications related to road network hierarchy per year was created (Fig.2).
Fig.2 clearly demonstrates that the evolution of the relevant literature is not stable. Over the past two decades, there have been several fluctuations. Local peaks and troughs can be observed, as evident from the irregular shape of the line in the corresponding diagram. Notably, the highest number of publications occurred in 2022, 2021, and 2019, with five relevant articles published each year. This was followed by 2014 and 2008, in which four publications were published each year. In the remaining years, the number of relevant articles ranged from one to two, with a few cases where three articles were published per year, acting as minor “outbreaks.”
Recently, there has been a steady increase in the relevant literature, indicating a dynamic trend toward future research. However, this trend is relatively new, as, prior to 2020, the corpus of literature experienced significant outbreaks, with a substantial number of articles published every five or six years. The cumulative diagram in Fig.2 portrays the cumulative dynamics of the literature evolution. As depicted in this figure, the literature evolution can be divided into two discrete stages. The first stage extended until 2008, during which the corpus had only one relevant paper, indicating limited scientific debate on the realization of road network hierarchy in urban areas. However, 2008 marked a turning point, after which the relevant literature expanded significantly, a trend that persists until the present. In the second stage, notable changes in the slope can be observed at four specific moments: 2011, 2014, and 2018. Regarding the publications on road network hierarchy, it is important to note that the sample includes a variety of scientific journals from different fields. Specifically, there are 32 journals representing 9 general fields and 18 specific fields. This diverse range of journals reflects the multidimensional nature of the road network hierarchy and its interdisciplinary nature. The literature covers various aspects, such as social, economic, and technological factors, going beyond the traditional perspectives of urban planning and transport. These journals, along with their fields and the number of relevant papers they contain, are presented in Tab.1.
It is worth noting the representation of different fields in the discussions around the urban road network hierarchy. Each journal is affiliated with a general field and a more specific field, determined through Scopus rankings. According to Tab.1, the most prevalent general fields are Social Sciences, with 15 papers (46.49%), and Engineering, with nine papers (28.13%). Another field with notable impact is Other Environmental Science, which had two papers (6.25%). Moving on to the prevailing specific fields, geography, planning and development ranks first with eight papers (25.00%), followed by urban studies with five papers (15.63%). Automotive engineering, general engineering, and mechanical engineering are tied in third place with two papers each (6.25%). The combination of prevailing fields and the overall variety of fields illustrates the interdisciplinary nature of the urban road network hierarchy. Therefore, policy recommendations should adopt integrated approaches rather than conventional, monothematic ones.
Turning to the scientific journals themselves, “Transportation Research Record,” a renowned journal in the field of Engineering, has the most relevant publications, with four papers. “IATSS Research,” representing the field of urban studies, follows with three publications. “Accident Analysis and Prevention,” “Sustainability,” “International Journal of Sustainable Transportation,” “Case Studies on Transport Policy,” and “Urban Design International” also make significant contributions, each with two publications. “Accident Analysis and Prevention” belongs to the field of medicine, while the others fall under the social sciences.
3.2 Content analysis
At this point, a content analysis of the findings deriving from the SLR process follows.
3.2.1 Development of the SLR overview table
The approach of SLR, particularly meta-synthesis (
De Gruyter et al., 2019), proves exceptionally beneficial. It utilizes a systematically structured methodology for the collection and analysis of relevant data. Moreover, this technique probes more profoundly into the core of each paper, emphasizing essential concepts and viewpoints. To facilitate this analysis and provide a comprehensive overview, an overview table (Table A1) has been created (Appendix A). This table includes detailed information about the articles in the sample, as well as future directions and insights into the urban road network hierarchy.
Table A1 serves as the main tool for analyzing and distilling meaningful knowledge from the articles. It consists of 13 columns, each displaying different information and features. The first column lists the authors of the articles, while the second column notes the publication dates. The third column highlights the key topic represented by each article, focusing on the constituent concepts of the urban road network hierarchy. Notably, only one topic is included for consistency reasons.
The fourth column presents the objective of each work, aiming to identify the role of the urban road network hierarchy within the context of the entire article. The subsequent two columns address the research methods used. The fifth column indicates the general research method employed (quantitative, qualitative, mixed), while the sixth column specifies the certain type of method utilized (methodological, case study, analysis, etc.). Each paper was assigned to a single category that was deemed most appropriate for that particular study, ensuring clarity and coherence.
The seventh column illustrates the spatial level at which each research work is focused (national, urban, route, etc.), and the eighth column specifies the exact study area, whether real or hypothetical. The ninth column provides a summary of the main findings of each article, while the tenth column presents any proposed methods for classifying the urban road network, if applicable. The eleventh column outlines the approach taken by each study in defining the urban road network hierarchy. This approach can be either conventional or alternative (multimodal), connecting each paper to the theory of urban road network hierarchy and the prevailing social and political goals.
The twelfth column indicates whether the proposed or existing road hierarchy system considers the urban dimension when formulating road categories. This finding highlights the potential connections between transport and the urban domain. Finally, the thirteenth column focuses on the future, highlighting the contribution of each study to the formulation of a new urban road network hierarchy system. The values in this column range from 1 to 5, where 1 represents a very low contribution and 5 represents a very high contribution. For the sake of clarity and readability, Table A1 is presented separately in the Appendix of this paper (Appendix A).
3.2.2 Descriptive analysis and general comments on the SLR table
Upon careful examination of Table A1 and subsequent descriptive statistical analysis, several noteworthy findings emerge. The majority of articles in the sample (59.52%) adopt an alternative approach to the urban road network hierarchy, indicating a discernible shift toward sustainability- and multimodality-centric perspectives within the literature. When it comes to the contribution of the sample to a new future hierarchy and the spatial level, Fig.3 is rather interesting.
Specifically, Fig.3(a) reveals that half of the papers exhibit a high or even very high level of contribution. These two categories are followed by papers characterized by a medium level of contribution (40.48%), while only 9.52% of the examined studies have a low or very low level of contribution. This finding signifies that current research related to road network hierarchy not only embraces an alternative rationale but also formulates effective strategies for future transport scenarios. As mentioned, another intriguing feature is the spatial level addressed by the papers included in the sample. As shown in Fig.3(b), more than half of the papers pertain to the urban or municipal spatial level, while a notable percentage (21.43%) focus on hierarchy strategies for specific routes. This indicates that the urban road network hierarchy is not confined to urban or metropolitan policy schemes on a macroscale; rather, it can be implemented on a microscale. Thus, this tool functions as a multilevel strategy in urban areas.
Fig.4 depicts the number of road/street categories suggested in the papers included in the sample. The number of road/street categories is particularly crucial, as an effective hierarchy system requires a proper balance (
Kosztolányi-Iván et al., 2019). Specifically, when a system has numerous categories (more than 10), achieving readability can be challenging. Conversely, a small number of categories (less than 5) fail to involve all potential aspects and priorities, thereby neglecting the various modes encountered in urban road environments. In general, a small number of categories corresponds to conventional approaches, while a large number aligns with alternative approaches.
A considerable percentage of the related studies (21.43%) do not propose a specific hierarchy method but instead they extensively examine existing hierarchy methods. These papers do not provide any relevant information about the number of street categories and are represented as “no data” in the diagram. The remaining articles displayed a range of 2 to 25 hierarchy categories. Notably, papers in five categories had the highest percentage (16.67%), followed by those in four categories (11.90%). This indicates that the current literature primarily employs urban road network systems with a small number of categories, suggesting a potential need for future alternative approaches with more concise hierarchy schemes. When looking into the “optimal” number of categories, i.e., above four and less than 11, it is evident that 42.86% of all the papers fall within this range (including those that do not propose a new hierarchy system). This signifies that a substantial proportion adopts “the right mix”, but there should be a greater focus on formulating alternative and innovative schemes with an appropriate number of categories.
What is more, the descriptive analysis of the SLR table identifies both general and specific research methods utilized. It is important to note that out of the 34 papers, only nine (21.43%) exclusively adopted qualitative research methods, while the rest were almost equally divided between those that utilized quantitative methods (43%) and those that employed mixed methods (quantitative and qualitative) (36%). This indicates a distinct trend in the literature to prefer quantitative tools and procedures.
Fig.5 provides further insight into the actual types of methods used, such as methodological, modeling, and review methods, which also shed light on how the urban road network hierarchy is analyzed, determined, and evaluated.
The figure (Fig.5) clearly demonstrates that more than a quarter of the sample is methodological in nature, followed by papers undertaking “experimental applications” with significant percentages (e.g., analyzing changes in road hierarchy). Similarly, articles incorporating analysis techniques (primarily with geographical criteria) to examine the current situation and propose future recommendations constitute a large portion of the literature. Therefore, papers related to urban road networks mainly fall into the categories of methodological, analytical, and experimental research, with fewer being conceptual, review-based, or case study-based.
3.2.3 Interpretating the SLR table: Trends and insights
The SLR table (Table A1) has contributed significantly to identifying current trends and providing valuable insights for future transport scenarios. One observed trend is the utilization of road hierarchies to recognize the morphology of urban road networks and highlight major road segments (
Jiang, 2009;
Huang et al., 2016;
Han et al., 2020). Similarly, Paraskevopoulos et al. (
2022) employed space syntax analysis to classify the road network in the Athens Metropolitan Area. Other similar research trends or ideas are reflected in the articles of Iván (
2014) and Kosztolányi-Iván et al. (
2019) who placed special emphasis on user participation, to determine which and how many road network categories a hierarchy system should have. Also, Chan and Cooper (
2019) investigated whether hierarchy can be used as a tool for predicting traffic flows.
Another research direction focuses on utilizing advanced algorithms, particularly machine learning techniques, for the automatic classification of urban road networks. Noori et al. (
2020), Bosurgi et al. (
2019), D’Andrea et al. (
2014), and Gülgen (
2014) demonstrated methodological approaches for determining the geographical hierarchy of road networks. Gharaee et al. (
2021) introduced a novel learning-based approach, adopting graph convolutional neural networks, to classify the road network in urban areas. Tsigdinos et al. (
2021a) used network analysis algorithms to introduce autonomous vehicles into a street classification system. However, it is important to note that these studies primarily follow a conventional, car-oriented approach, which raises questions about their applicability within a sustainability-oriented framework in the future.
Moving on to different research directions that emphasize the social dimension and multimodality of streets and roads, it is important to highlight several articles that aim to redefine the hierarchy of urban road networks through the promotion and adoption of “complete streets” policies.
In particular, Gregg and Hess (
2019) conducted a comprehensive literature review on the design issues of complete streets, while Elias (
2011) gathered and categorized the appropriate criteria for designing such streets. Additionally, Kingsbury et al. (
2011), Hui et al. (
2018), and Kumar et al. (
2019) developed innovative tools to assess the quality of these schemes and related infrastructure. Consistent with this, Dehghanmongabadi and Hoskara (
2022) created a holistic method outlining the main steps for planning an efficient complete street, considering both the built environment and the social context.
Delbosc et al. (
2018) contributed to the scientific debate by comparing complete streets (predominantly favored in the USA) with the corresponding policy scheme in Australia (smart roads). Their conclusion highlighted the benefits of a combined rationale as a means of introducing a new and alternative road hierarchy approach. Similarly, Curtis and Tiwari (
2008) and Tsigdinos et al. (
2021b) highlighted the value of multimodal corridors and their positive impact on the urban road environment. They proposed cohesive strategies for redesigning urban arterial roads. Furthermore, Mirzahossein et al. (
2022) made a significant contribution by outlining a cohesive method to prioritize arterial roads for transformation into multimodal corridors. Another take from Sheikh-Mohammad-Zadeh et al. (
2022) formulated an assessment process based on transit, access, and place, which could serve as the basis for a new classification system.
Other works of significant interest and contribution to the formulation of a new urban road network hierarchy are those of Stamatiadis et al. (
2017), Tsigdinos and Vlastos (
2021), and Tsigdinos et al. (
2020). These works established a coherent methodological framework for redefining the hierarchy of the urban network at the metropolitan and citywide levels utilizing the two-dimensional classification approach initially proposed by Jones et al. (
2008) and Jones and Boujenko (
2009) through the concept of Link and Place. This concept, which is steadily gaining popularity, combines the urban planning and transport dimensions, taking into account the multimodality of urban streets and the variety of activities that occur within them. A similar rationale was adopted by Oguchi (
2008), who proposed a two-dimensional matrix for formulating road hierarchies at the national level in Japan. Additionally, an interesting study incorporating this bidimensional classification was conducted by Su et al. (
2022), who proposed an activity-based street reconfiguration scheme, thereby expanding the scope of traffic-related road hierarchies.
Finally, several works embrace the social dimension of streets, effectively linking the concept of hierarchy with similar concepts and serving as a catalyst for future studies that move beyond conventional practices. For instance, Charlton et al. (
2010), Walker et al. (
2013), Charlton and Starkey (
2017), and Theeuwes (
2021) examine the connection between the urban road network hierarchy and road safety, while Carmona (
2015) and Sauter and Huettenmoser (
2008) explore hierarchy in relation to land use/activity and social relations. These perspectives demonstrate that the urban road network hierarchy is not merely a process of categorizing streets but rather a broader framework that influences other aspects of urban street environments and social nexus.
3.2.4 Identifying the fundamental components and their evolution
In addition to descriptive statistics and significant trends or insights, the SLR table facilitated the identification of the fundamental components of the urban road network hierarchy. These components are sub-concepts that are closely intertwined with the main concept of hierarchy. This type of analysis is also evident in relevant literature review papers (
e.g., Lagorio et al., 2016) and serves as a valuable tool for gaining a deeper understanding of the main concept, i.e., the urban road network hierarchy.
These components essentially represent the main focus of each article included in Table A1. They were determined through the careful interpretation of the papers in the SLR sample. This process distinguished 16 relevant constituent concepts that span a broad spectrum of the urban and transport domains. These concepts were thoughtfully selected to ensure coherence and comprehension, aiming to capture the general content. All these components are presented in Fig.6 and are organized according to specific time periods. These periods were established based on the number of components and the year they were first introduced, resulting in a clear depiction of their evolution over time. Notably, the evolution of these components serves as a meaningful indicator for contextualizing the urban road network hierarchy and the underlying process. In Fig.6, the components introduced for the first time are depicted in gray, while those that reappear are shown in light purple.
During the 2000–2010 period, seven new components emerged (found in eight articles). The concept of Link and Place, which pioneered two-dimensional classification, is featured in two publications. Therefore, alternative (more sustainable) approaches to hierarchy have been part of the literature since then.
During 2011–2015, ten new scientific articles introduced eight novel concepts, with road safety being the only concept adopted from previous research. In contrast, previously prevalent concepts such as network analysis, street reconfiguration, and multimodality disappeared. This shift exemplifies the dynamic and diverse nature of the urban road network hierarchy, where concepts evolve and others wane. Notably, ‘complete streets’ and ‘pattern recognition’ emerged as dominant concepts during this period, reflecting interdisciplinary influences from urban mobility and computer science (image processing), respectively.
In the subsequent period, 2016–2022, the field stabilized, with 23 new publications. Of the eight concepts discussed, only two—trial classification and spatial analysis—are new. The remaining concepts re-emerge from earlier periods. The dominant themes included network analysis, street reconfiguration, and complete streets, followed by multimodality and pattern recognition. This finding indicates a balance of qualitative and quantitative research orientation in the literature.
These three stages involve shifting dominant concepts, likely influenced by advancements in technology, computational methods, urban planning, transport priorities, and community awareness. For example, pattern recognition is technology-driven, whereas complete streets stem from human-centric priorities. Understanding these evolving concepts is not a mere identification task; it requires additional analytical tools. Tab.2 elucidates the relationships between various concepts and the methods employed in their study.
Tab.2 enhances our understanding of the methods utilized in each fundamental concept. Generally, the range of methods is evident, as almost every concept, that is represented by multiple papers, has been examined using different approaches. When examining the prevailing concepts closely, the following points should be highlighted: On the one hand, it appears that certain concepts are investigated primarily using specific types of methods. For example, “complete streets” are predominantly discussed in review papers, “network analysis” is frequently found in analysis papers, “street reconfiguration” is principally encountered in methodological articles, and “traffic safety” is found in studies carrying out experiments. On the other hand, concepts such as “multimodality” and “pattern recognition” are found in a variety of paper types, indicating greater diversity.
In summary, the evolution of these components highlights two key aspects: a) the “diverse” nature of the hierarchy and b) the continuous enrichment and expansion of the literature with new constituent concepts that enhance coherence while redefining the concept of hierarchy itself over time. Furthermore, certain concepts are closely associated with specific methods, while others are amenable to various tools and techniques.
3.2.5 Citation network analysis
Citation network analysis is an integral part of SLR and has the ability to shed light on the interactions among scientific articles through an extensive network of citations (
Lagorio et al., 2016;
Aria and Cuccurullo, 2017). In this study, citation network analysis was employed to visualize the connections among the literature related to urban road hierarchy. Only the papers included in the sample were considered, in line with Lagorio et al. (
2016) and Gustafsson et al. (
2014), to ensure feasibility and facilitate interpretation. This tool helps identify influential articles and those that lack research interaction. It is based on graph theory, widely utilized in network analysis (
Rodrigue, 2020), and proves particularly valuable for recognizing interdependencies and relationships. Specifically, articles are represented as nodes, and citations are represented as edges, forming networks with specific topological properties (
McLaren and Bruner, 2022). This study adheres to the convention where each arrow points from the citing document to the cited document, representing the “cite” relationship.
According to Fig.7, our citation network consists of three main entities, indicating that the literature on urban network hierarchy is influenced by various factors. First, there are several articles that have no contact or relationship with the rest of the relevant literature; these articles form 11 sub-networks consisting of only one node (blue color in Fig.7). Notably, these 11 articles, the majority of which were published after 2010, appear to be isolated without any interaction with the main body of literature.
The second entity consists of six articles, forming a small path with six nodes and five edges (green color in Fig.7). At this point, we utilize the connectivity index γ, defined as follows (
Rodrigue, 2020):
where e are the edges of the graph and v are the vertices of the graph.
The γ index of this small path, which equals 0.42, shows that the graph is not particularly connected. Consequently, as evidenced by its linear shape, the articles do not interact with each other, except for two of them (
Charlton et al., 2010;
Charlton and Starkey, 2017), which have node degrees of 3 and 2, respectively. These nodes are more connected, reflecting the prominence of these two projects in the entire sub-network. Additionally, it should be noted that this path is not connected to the main network and functions as a separate entity. This sub-network first appeared in 2010, and the articles belonging to this cluster primarily address road safety and pattern recognition issues.
Finally, the main body of literature establishes a large sub-network of articles and citations characterized by a complex shape (red color in Fig.7). This sub-network clearly demonstrated the interactions among most of the articles included in the sample. It consists of 25 nodes and 42 edges. The connectivity index γ, which is 0.61, indicates that the current sub-network is moderately connected. This implies that some of the articles (nodes) consider previous ones, and in general, the influence of older research works on newer ones can be considered relatively significant. The articles included in this sub-network cover a wide range of topics related to urban road network hierarchy, such as geography and development, urban studies, transport, computer science, and environmental sciences. These conditions highlight once again that hierarchy is not an isolated issue but resides within a wider scientific debate. Moreover, the interaction of these different types of articles and topics cultivates a comprehensive context for understanding the urban road network hierarchy.
In accordance with Fig.8, which was generated using yEd software, known for its network management capabilities, the most influential articles based on their citation count within the corpus are as follows:
First, the studies conducted by Jiang (
2009) and Jones et al. (
2008) are notable, with a node degree of 6. These works have evidently influenced all the connected articles, demonstrating their critical value to the entire literature. These papers can be regarded as the principal pioneers of the relevant literature. Next, the research of Jones and Boujenko (
2009) and Stamatiadis et al. (
2017) followed with a node degree of 5. Finally, notable studies that have influenced the literature corpus include Huang et al. (
2016), Kingsbury et al. (
2011), and Charlton et al. (
2010).
To summarize, the articles highlighted by the SLR process appear to represent the most crucial research on the urban road network hierarchy in the literature corpus. This process not only illustrates the evolution of the literature but also demonstrates the profound and continuous interactions among its constituent parts, i.e., the articles. Furthermore, the citation network analysis clearly revealed which articles play a vital role in shaping the complex landscape of the literature corpus.
4 Discussion
4.1 Main findings and research issues
In a rapidly evolving research landscape in the field of urban and regional design (
Cao and Zhang, 2022), our research dealt with a fundamental notion of both urban and transport planning: the urban road network hierarchy. This work is one of the few to date that scrutinized the complex process of classifying streets and roads from multiple perspectives. The primary objective was to contextualize the notion of hierarchy and determine its properties. By employing a SLR process, this research sheds light on existing trends and the evolution of relevant literature over time. This study provided a comprehensive overview of the selected sample, including its composition, characteristics, and development of the related literature. The use of this review method in previous urban and transport studies (
e.g., Scheepers et al., 2014;
Wang et al., 2022;
McKane and Hess, 2023;
Papaix et al., 2023) further enhances the value of our study.
Our analysis revealed a significant increase in the number of articles published primarily after 2010, with peak values occurring in 2019, 2021, and 2022. The journals included in our sample represent diverse scientific fields, reflecting the multidimensional nature of the urban road network hierarchy. However, it is noteworthy that the fields of geography and urban studies, followed by engineering, are the most prominent. Despite this spatial-centric focus, road hierarchy remains an interdisciplinary topic requiring integrated approaches for analysis and transformation toward sustainability. Pumain (
2006) and Corominas-Murtra et al. (
2013) also discussed the inherent nature of hierarchy in depth. In addition to general descriptive bibliometric analysis, our study developed an overview table using the SLR methodology to provide detailed information on the included articles and offer significant insights into the urban road network hierarchy (similar tables can be found in
Butler et al., 2021). This table consisted of 13 columns, including information such as the objectives, methodology, proposed hierarchy systems, and key findings of each article. Through this comprehensive overview, readers can gain a clear understanding of the characteristics of the literature in this field. Furthermore, this table helps identify emerging trends and dynamics toward future transport scenarios.
The descriptive statistical analysis of the table features provides valuable insights. First, it was demonstrated that the majority of articles focus on the urban and municipal spatial level, while others address specific routes. This indicates that the urban road network hierarchy is not limited by macroscopic and strategic analysis but can also be implemented at the microscale, thereby serving as a multilevel strategy in urban areas. This finding is in line with that of Okraszewska et al. (
2018), who emphasized the significance of a multiscale transport system planning framework.
When it comes to the methods employed, most articles utilize mixed and quantitative methods (
e.g., Hsu and Lin, 2011;
Dong et al., 2013;
De Baets et al., 2014), indicating a clear preference for spatial analysis or geoinformatics either alone or coupled with organizational procedures to determine road hierarchy schemes. The use of mixed procedures represents a recent trend in hierarchy analysis, combining the development of new categories with the determination of their geographic attributes (also see
Rychlewski, 2016). With regard to the approaches utilized (conventional or alternative), it is encouraging to observe that the alternative approach has a notable influence on the majority of published works, emphasizing sustainable modes of transport and multimodality. This shift aligns with the narratives surrounding sustainable mobility (
Holden et al., 2020) and aims to promote a new ethos of mobility.
In addition to the adopted approach, it is important to highlight the proposed hierarchy systems. Categories with 4 or 5 levels are the most prevalent, while approximately 40% of the papers fall within the optimal range of 5 to 10 categories. This indicates that the current literature tends to employ urban road network systems with a limited number of categories, presenting a somewhat mixed message for the future-advocating for an alternative approach but with more concise hierarchy schemes. Furthermore, 55% of the papers considered the urban planning dimension, creating a conducive environment for the promotion of sustainable mobility schemes. Existing literature on transport planning embraces integrated approaches (
Miller, 2018;
Capasso Da Silva et al., 2020), aligning with the current research trend of adopting cohesive strategies that include both urban and transport dimensions (
Sarri et al., 2023). Finally, most papers make a substantial or even highly significant contribution to the establishment of a new hierarchy. This demonstrates that current research on road network hierarchies not only adopts an alternative rationale but also strives for coherent strategies in future transport scenarios.
Furthermore, the meta-synthesis process revealed four main research trends pertaining to the hierarchy of urban road networks within the sample. The first trend focuses on analyzing the morphology and structure of urban road networks, while the second trend aims to develop techniques for automatically determining the hierarchy of road networks. In contrast, the third and fourth trends underline the social dimension of streets, proposing multidimensional hierarchy systems that include urban planning and transport aspects. Despite their differences, these trends should be seen as complementary rather than conflicting, as both spatial and social perspectives are essential in understanding road networks. Integrated approaches, such as those discussed in Melkonyan et al. (
2020), advocate for a holistic view of urban and transport systems, utilizing both analytical tools and organizational strategies. Moreover, given the uncertainty of the future, it is crucial to adopt inclusive and multidimensional approaches rather than narrow approaches to envision and shape desirable future scenarios (
Lyons and Davidson, 2016;
Vallet et al., 2020). As a result, these four emerging trends can be consolidated into two overarching categories, namely, analysis and planning, which can serve as driving forces for future transport street classification planning schemes.
Shifting focus to the contextualization process, this study distinguished the components closely associated with the hierarchy of urban road networks, specifically the concepts that are most relevant in this regard. The SLR summary table indicates a total of 16 concepts that are interconnected with various aspects of urban environments. As depicted in Fig.9, the prevailing concepts cover a wide range of dimensions, including “complete streets,” “network analysis,” and “street reconfiguration,” all of which are closely linked to hierarchy. Additionally, there are concepts that may not initially appear to be highly relevant; however, they significantly impact the road hierarchy within a city (e.g., pattern recognition). The analysis of these components reveals two key insights: the diverse nature of hierarchy and the continuous emergence of new constituent concepts within the literature. The evolution of these concepts can be attributed to advances in technology, improvements in computational methods, shifts in urban planning and transport priorities, and increased community awareness.
Another important aspect to consider is the analysis of the citation network. This process identified three main entities or subnetworks:
a) Independent studies that do not cite each other or the rest of the sample.
b) A small path consisting of a limited number of papers that are connected to each other but not to the rest of the literature.
c) The main body of the published corpus, which is a dynamically connected entity (with an adequate value of index γ) that demonstrates the interaction between the papers in the sample.
In particular, the third sub-network constitutes the driving force behind new publications covering a wide spectrum of topics related to urban road network hierarchy, such as geography, urban studies, engineering, transport, computer science, and environmental sciences. The analysis of the citation network provides insights into the evolution of the literature and reveals the continuous interactions between the articles. Furthermore, we explored which articles had the most influence on the development of the literature corpus.
When it comes to the research questions, the SLR and supplementary methods such as citation network analysis have provided insights for each research question. Specifically, in response to the question about the nature of hierarchy (question a), the study has demonstrated that hierarchy is a concept that enables a comprehensive understanding of urban space and is an essential tool for future cities. It plays a central role in spatial planning, including both urban and transport dimensions. This study highlights the critical role of hierarchy, which should be further enhanced in the future.
Concerning whether hierarchy is mostly qualitative or quantitative (question b), it has been shown that hierarchy is not solely one of these options but rather a dynamic combination that involves both qualitative and quantitative approaches.
With the regards to the question of whether hierarchy emerges or is solely a matter of planning (question c), the findings indicate that hierarchy sometimes emerges from the topological properties of space. However, it has been proven to be the result of both planning and auto-formulation.
In general, the urban road network hierarchy should be deemed as a constitutional element of all (strategic) urban and transport planning schemes, regardless of spatial scale. Other planning and design activities should be guided and controlled to ensure a cohesive and coordinated transformation of the urban environment. Within this framework, a new definition considering multimodality and sustainability is articulated. Specifically, “road network hierarchy is the process of categorizing roads based on at least two features, (e.g., urban activity and transport significance) with the aim to prioritize the movement of sustainable modes through the lens of an integrated and multimodal analysis and planning perspective.” Hence, this new definition embraces an alternative approach that now serves as the cornerstone of an integrated planning process. It places importance on sustainable modes of transport, such as active modes and collective transport, while also considering multimodality or coexistence by proposing concepts such as complete streets, shared spaces and other similar ones. The proposed definition aims to facilitate sustainable and viable transport futures, where the road network hierarchy plays a crucial role in overseeing both urban and transport planning procedures and could function as a mechanism for coordinating different transport modes in accordance with the principles of sustainability.
Specifically, the uncertainty and complexity of transport futures have been widely recognized (
Vallet et al., 2020). With the emergence of new modes, technologies, and measures, it is crucial to avoid dealing with these modes in isolation to prevent chaos and inequities in future cities, particularly in the domain of transport. Consequently, road network hierarchy should be prioritized and it should prevail as a centerpiece of future planning rationales. A planning rationale that incorporates both urban and transport planning will organize and categorize all these new modes and measures under a common “umbrella.” This approach ensures sustainable transport futures where technological advancements, people, and the environment can coexist harmoniously, operating and living in a coordinated way. The role of hierarchy, in this case, becomes that of a “binder” for cities, fostering cooperation and coordination. Thus, the emphasis should be on adopting integrated approaches to avoid conflict and the creation of dysfunctional and dystopian futures caused by antagonistic new modes. Consequently, the interpretation of roads as corridors of movement and stay expanding beyond the ground floor (air, sea, underground) and as an extent the concept of road network hierarchy as the general way of classifying every transport network in urban areas are actions directly linked to sustainable transport futures. Interestingly, sustainable transport futures may view a sustainably oriented hierarchy as a prerequisite. In this context, hierarchy aspires to inaugurate a planning rationale through understanding strategic perspectives and providing guidance to lower spatial levels. It functions as a means to radically rethink urban environments while also serving as a tool for reimagining and shaping urban futures toward sustainability. This article serves as a facilitator for this shift from conventional rationales to alternative and sustainable ones.
4.2 Contribution
This paper sheds light on the significance of integrated transport and urban planning by examining and breaking down one of its fundamental planning tools, namely, the urban road network hierarchy. This is one of the first studies to comprehensively explore the road network hierarchy from a broad perspective using a systematic and holistic approach. More specifically, the SLR process revealed established or emerging concepts (e.g., multimodality, link and place, network analysis) and research trends (e.g., focus on the social dimension of streets), thus enriching the literature. Furthermore, this study introduces a new definition for road network hierarchy, establishing a direct connection to sustainable transport futures.
In this context, the research presented herein has the potential to significantly contribute to the scientific discourse surrounding the redefinition of hierarchy in urban planning. Specifically, it offers insights into the integration of multimodality and sustainability, which can serve as a catalyst for the development of a new mobility culture that prioritizes people over cars. In essence, by comprehending road hierarchy as a comprehensive planning tool and a facilitator of alternative modes of transport, it is anticipated that a paradigm shift toward coexistence and inclusivity can be achieved (
Holden et al., 2020). Moreover, the findings of this study have implications for policy-making and practical implementation, as they can guide consultants, policymakers, and planners seeking to adopt integrated planning approaches and effectively utilize the concept of the urban road network hierarchy. It is worth noting that the road network hierarchy plays a fundamental role in establishing effective urban transport management strategies, guiding the implementation of sustainable policies and interventions such as travel demand management and promoting active modes of transport (
Ling et al., 2021).
Hence, policymakers and administrators stand to benefit on two fronts. First, a revised road network hierarchy was established that gives precedence to sustainable modes of transport through spatial configuration (e.g., the designation of ring roads for motorized vehicles and radial arterial roads for public transport and active modes), as well as the creation of new categories specifically designed to accommodate these modes (e.g., the inclusion of active mode boulevards or strategic public transport corridors). Second, policy measures, interventions, and maintenance activities that adhere to the established hierarchy should be implemented. For example, active transport boulevards should be equipped with dedicated cycling infrastructure and sidewalks with a minimum width of 2.1 m.
4.3 Limitations and future research
Understanding and contextualizing the hierarchy of urban road networks is a complex issue that cannot be fully explored by a single study. Continuous effort in undertaking further research initiatives is essential to enrich the literature and influence the planning practices of local and international communities. In this framework, the SLR process documented the components of the urban road network hierarchy and their evolution over time. However, it did not proceed to analyze the relationship between these factors in detail, as such an objective would require different explanatory tools. To this end, future research might examine how these constituent elements interact with each other through structural analysis or participatory multicriteria techniques. Conducting in-depth interviews with various experts, such as policy makers, stakeholders, consultants, politicians, and associations, would be beneficial in this context. Additionally, while the SLR process aimed to encapsulate the most relevant articles; therefore, the number of the final pool may have appeared small. However, in the context of a broader work addressing urban road network hierarchy (including also gray literature), more papers could be useful. What is more, this research concentrated on the forms of urban road network hierarchy, rather than the definitions within each scientific paper. Hence, new papers could shed light on the content of urban road hierarchy, thus articulating new definitions and interpretations.
Regarding further research options, it is recommended that new efforts could concentrate on developing an analytical policy framework related to different road hierarchy schemes. This framework is crucial for facilitating the implementation of road hierarchy principles in real-life scenarios, effectively bridging the gap between theory and practice. The development of this analytical framework could result from a comprehensive qualitative approach that incorporates both systematic reviews and interviews.
Apart from studying hierarchy itself, this study highlights the importance of examining the impacts of hierarchy on accessibility and equity. For instance, Liao and van Wee (
2017) demonstrated how different transport hierarchy options, such as car-oriented or public transport-oriented systems, can affect the robustness of the transport network through accessibility measures. Another related study conducted by Liao et al. (
2017a) showed how various transport scenarios, each with different hierarchy priorities, can influence accessibility and travel patterns. In this context further research could be carried out, considering the appropriate social and demographic indicators to detect how new road hierarchy systems influence the opportunities of people to move and interact in urban areas. Evaluating hierarchies using multiple indicators is crucial for future research. Notably, one potential area of investigation is how road network hierarchy can contribute to ensuring the resilience of transport systems, as highlighted by similar research conducted by Ma et al. (
2023) on urban road network configuration and resilience. Furthermore, future research uptakes should explore the relationship between road management and hierarchy in more depth, especially with regard to safety issues (
e.g., Liu et al., 2023) and traffic management (
e.g., Sirmatel and Yildirimoḡlu, 2023).
Finally, beyond urban road network hierarchy related to the “ground floor,” future research works can also focus on the organization of other transport modes, such as air, sea, and public transport, particularly metro or rail systems operating underground. This idea emerged during the preparation of this paper and presents an exciting challenge for further research. Ultimately, hierarchy and classification are fundamental concepts in all transport systems, whether rural or urban. Tellingly, further research is necessary to advance our understanding of road network hierarchies. Exploring new research questions and topics will contribute to the scientific literature on this subject and inform daily practice, promoting accessibility, energy efficiency, and overall sustainable cities.
5 Conclusions
This paper covered a critical topic that underpins sustainable city planning—the urban road network hierarchy. The key insight is that this hierarchy goes beyond being just a tool for transport planning and should be considered a way to rethink the urban environment. It is argued that hierarchy is a fundamental concept that can shape an integrated urban and transport planning framework for the future. It may function as the cornerstone of new spatial planning schemes with the aim to coordinate the complex nature of the urban domain. Due to its multiple dimensions, the examination of hierarchy requires both qualitative and quantitative methods. Various research trends have emerged, offering different possibilities and opportunities for further studies.
Moreover, when planning hierarchy systems, it is important for policy makers, planners, and consultants involved in the process to exercise caution, as hierarchy can impact every aspect of the realities of urban areas and transport in cities. This research provides planners and stakeholders with valuable knowledge that can assist in redefining the road network hierarchy in their respective areas of responsibility, particularly in terms of facilitating sustainable modes of transport. For instance, new categories could be formulated to prioritize active modes, or advanced geospatial tools could be developed to configure the new strategic road network.
The main tool for contextualizing the notion of urban road network hierarchy was our SLR. By building an organized framework of collection, categorization and analysis of relevant papers, we explored in depth and underlined critical issues such as the components and evolution of the existing body of literature. Overall, four main research trends were identified: a) road morphology and structure, b) advanced algorithms for street classification, c) integrated street classification planning, and d) the social dimension of street classification.
This paper contributes to the current state of the art by highlighting the interdisciplinary nature of road hierarchy, emphasizing that it is more than just a planning tool. It offers insights into how the transport network functions and its impact on other dimensions, such as safety, sustainability, and multimodality. In addition to its theoretical value, this study intends to influence the planning process itself by underscoring new attributes of hierarchy that may have gone unnoticed. Future studies or projects can benefit greatly from this scene-setting work, ultimately leading to the creation of just and sustainable cities.
The Author(s). This article is published with open access at link.springer.com and journal.hep.com.cn
PDF
(2178KB)
