3.1.1 Time and Geography

The literature on comfort in active mobility has grown steadily over the past decade. The topic gained momentum in 2018 and peaked in 2022 with ten papers published about comfort in active modes with a minor drop in 2021 during the COVID pandemic. The first papers past our selection criteria were published in 2012 (Figure 2A). Except for one paper that studies walking and cycling comfort together, comfort in these modes has always been studied separately, and cycling constitutes the majority of our target selection (33 papers out of 58). Geographically, the research is led by studies from the United States (n = 10, 17%) and China (n = 9, 15%), followed by Canada (n = 7, 13%), Sweden (n = 4, 8%) and several European countries, though contributions are globally distributed (Figure 2B).

Figure 2: A) Temporal trend B) Geographical distribution.

3.1.2 Research Field and Authors

Besides transport, active mobility comfort has been studied in other research fields such as urban planning, health, safety and accidents, environmental studies, and computer science, highlighting its interdisciplinary nature. It is noteworthy that more than 30% of the studies were published in 4 journals (Transportation Research Record, Transportation Research Part F, Transportation Research Part A and Accident Analysis and Prevention). While many of the cycling papers are published in transportation and accident/safety journals, walking comfort has been discussed more diversely in urban planning, environmental studies and computer science. A total of 207 unique authors contributed to our collection of papers, but only 10% have published more than once about active mobility comfort. This indicates that “comfort in active mobility” is a niche and emerging topic without a well-defined core group of dedicated experts.

3.1.3 Data Collection and Analysis Methods

Thematic analysis reveals a strong concentration on “human perception and comfort,” particularly in cycling-focused studies, followed by research on environmental conditions and behavioral aspects, the latter being more prevalent in walking-related research. 58% of the studies (n = 34) used surveys and questionnaires as the data collection method, followed by instrumentation-based approaches (n = 15, 25%) and third-party data sources (n = 14, 24%). A smaller number of studies employ simulations (n = 11, 18%), digital platforms and apps (n = 9, 15%), and video recordings (n = 6, 10%), while interviews and observations remain marginal (n = 5, 8%). Almost half of the studies benefited from a mixed data collection approach (Figure 3A). Objective approaches prevail overall, although a notable share of studies integrate both subjective and objective measures. Quantitative approaches dominate the analytical landscape in both modes, with statistical analysis and modeling constituting the most applied method (n = 49, 84%), particularly in cycling-related studies. Other methods, such as GIS-based spatial analysis and literature analysis appear less frequently (n = 9, 15%) and are often tied to walking studies (Figure 3B). It is noteworthy that none of the studies in our review used qualitative analysis methods.

Figure 3: A) Data collection methods B) Data analysis methods.

3.1.4 Target Groups and Research Frameworks

Most research targets the general public (n = 29, 50%), with fewer studies addressing vulnerable populations (n = 9, 15%), location-based (n = 5, 9%) or skill-based groups (n = 5, 9%). Sample sizes vary, but small (1–50) and large (251–1,000) groups are most common (46% together). A considerable number of papers (n = 16, 27%) do not disclose sample size, which raises concerns about methodological transparency. Regarding theoretical orientation, 70% of the studies (n = 41) adopt an analytical framework, aimed at empirically evaluating comfort indicators or modeling influencing variables. Conceptual frameworks constitute a smaller portion of the papers (n = 13, 22%), often to map comfort-related constructs or design evaluation metrics. Only a handful of papers (n = 4, 6%) use theoretical frameworks such as bicycle level of service, which indicates low engagement with abstract or critical perspectives in the field.

3.3.1 Comfort as safety

In articles focused on interactions between pedestrians and other road users, comfort is often examined as safety (Bigazzi et al. 2021; Gill et al. 2022; Lubbe & Rosén 2014). Lubbe & Rosén (2014) see pedestrians as a possible trigger of pushing the boundaries of drivers’ comfort towards discomfort and lost control, the latter clearly separated by a sharp safety boundary (Lubbe & Rosén 2014). While focusing on drivers’ comfort, the studies show that safety and comfort are closely interrelated, but still different in detail. Mostly through video observation of road user interactions in parallel with subjective questionnaires, those interactions are shown as carrying a safety and comfort cost (Gill et al. 2022; Lubbe & Rosén 2014). Pedestrian comfort, per se, is clearly increased when vehicle users drive at slow paces, and safety perception is increased when vehicles visibly reduce their speed, “possibly because it indicates drivers’ awareness and intention” (Gill et al. 2022: 181).

A similar conflation of comfort and safety appears even more prominently in cycling research. In cycling pools, safety concerns together with comfort indexes make up more than half of the body of literature. Safety is the prominent topic when studying comfort, and many researchers have expanded the scope of safety measures to comfort studies. Several works examine physical separation between cyclists and motor vehicles, concluding that such infrastructure reduces discomfort, but their operationalization of comfort closely mirrors traditional safety logic. For example, Li et al. (2012a) and Li et al. (2012b) find that environmental separation significantly influences perceived comfort, based on a questionnaire study. Similarly, McNeil et al. 2015) show that physical buffers, even as minimal as plastic flex posts, substantially increase comfort ratings, particularly among the “interested but concerned” demographic. Studies investigating high-risk scenarios like bicycle-vehicle interactions (Apasnore et al. 2017), presence of trucks near loading zones (Abadi et al. 2019) and user-rated intersection designs (Monsere et al. 2020) also equate reduced exposure to vehicle conflict with enhanced comfort, without directly assessing how cyclists themselves define or describe comfort in non-safety terms.

A similar pattern emerges in more dynamic or psychologically focused studies. The Cycling Anger Scale (O’Hern et al. 2019) investigates discomfort through emotional reactions like frustration or aggression, yet its focus remains on external stressors such as cars and police presence. Experiments exploring obstacle avoidance maneuvers (Lee et al. 2020) and cycling under poor weather or road surface conditions (Shoman et al. 2023) both frame comfort through behavioral and physiological responses, linking it implicitly to risk rather than ease. Roundabout studies (Singleton & Poudel 2023) again highlight protective design as the driver of comfort ratings, reinforcing the overlap with safety perception. Perhaps the clearest case of conceptual conflation appears in Kazemzadeh’s work (Kazemzadeh et al. 2024), which examines comfort in shared spaces. The study employs BLOS and safety-inspired measures, such as passing distance and visibility, but concludes with claims about comfort, despite measuring indicators aligned with spatial safety. Through these studies, comfort is consistently approached as a function of infrastructural risk and protective features, with limited engagement with subjective, bodily, or experiential dimensions of comfort as perceived by cyclists themselves.

3.3.2 Dynamic/surface comfort

Road design is often addressed as a factor influencing the perception of comfort (Garilli et al. 2023; Huang et al. 2024; Piao & Lee 2022; Zuo et al. 2021), but only a few studies investigate it further. Wojnowska-Heciak et al. (2022) investigate the influence of different pavement types on the moving comfort of pedestrians with motor disabilities. The authors emphasize that poor-quality pavement surfaces are a common obstacle and that large, smooth concrete slabs are particularly preferred. Closely related to pavement is vibration comfort. For wheeling, the pathway roughness indices (PRI), adapted from the International Roughness Index (IRI) used for roadways, are employed to categorize whole-body vibrations experienced by wheelchair users (Duvall et al. 2016). The authors applied a physical gadget, called PathMeT, which measures the vibrations on different surface types (Duvall et al. 2016). The international standard (ISO 2631 – Mechanical Vibration and Shock, Evaluation of Human Exposure to Whole-Body Vibrations) is used as a base to evaluate this type of comfort. Compared to cycling, surface and dynamic comfort remain considerably less studied in walking research, with only a handful of studies addressing it directly.

In cycling literature, almost one third of the papers are focused on human-machine mechanisms. Sports engineering and applied sciences study rider comfort with regards to bike frame stiffness (Shah et al. 2020) and biomechanical sitting comfort (Scholler et al. 2021). They investigate the materials and structural design of a bike to ensure higher riding comfort for the average person using simulations and mathematical calculations and laboratory tests. Vibration comfort is another prominent theme in cycling comfort that refers to the level of vibration transmitted to humans on various surfaces, also referred to as dynamic comfort during the riding movement (Lépine et al. 2014; Vanwalleghem et al. 2013). There are certain standards introduced by the Standard Organization (ISO) that are investigated through laboratories and real-life situations. This kind of comfort mainly impacts cyclists’ choice of bicycle (Gao et al. 2018; Gao et al. 2019) and is the most defined type. Recent studies from a transportation engineering point of view investigate pavement conditions and humans’ perception of comfort using eye-tracking sensors, simulations, accelerometers and smartphone applications combined with subjective experiments and tests (Gadsby et al. 2022; Niska et al. 2024; Yang et al. 2023). This line of research gives insight into bike and road surface material and design.

3.3.3 Comfort indices

Recent efforts to quantify the pedestrian experience have culminated in the development of comprehensive assessment tools, most notably the Pedestrian Walkability and Comfort Index (ICCP) (Albala & Mülfarth 2024) and the Walkability Index by Al Shammas & Escobar (2019), later extended into a comfort-based routing algorithm (Al Shammas et al. 2023). By centering comfort, these indices must account for an expansive array of variables. The ICCP integrates thermal, acoustic, and ergonomic comfort dimensions alongside urban form and climatic attributes, though the authors acknowledge that visual, illumination, and air quality factors remain outside its scope (Albala & Mülfarth 2024). Al Shammas and colleagues (2023) take a more user-centered approach, allowing pedestrians to input individual preferences and generating the shortest comfort-based walking route within an interactive tool. Together, these approaches represent a shift toward integrating comfort into GIS-based walkability assessments, a task previously avoided due to the substantial computational effort required to model highly dynamic factors and diverse individual perceptions.

The indicators feeding these indices fall into two areas. The first concerns street-level urban design and infrastructure: Landscape layout and architectural function (Nury & Haykal 2023; Piao & Lee 2022; Ren et al. 2023; Zuo et al. 2021), separation of transport types (Zuo et al., 2021) and specific sidewalk characteristics such as pavement conditions and effective walking width (Duvall et al. 2016; Huang et al. 2024; Wojnowska-Heciak et al. 2022), and the presence of street furniture and general urban design amenities (Huang et al. 2024; Alsina-Pagès et al. 2024). The second concerns environmental and microclimate interactions: the material composition of building facades directly impacts thermal comfort (Takebayashi 2016), while street planting and greenery provide a “soothing effect” on both urban heat (Teshnehdel et al. 2020) and sound levels (Ren et al. 2023). Piao & Lee (2022) as well as Teshnehdel et al. (2020) argue that these tangible elements interact inseparably with intangible factors such as seasonal characteristics, natural lighting, and radiation, reflecting the inherently dynamic and multisensory nature of pedestrian comfort.

A parallel effort to develop standardized comfort metrics is visible in cycling research, though with a notably different disciplinary emphasis.

Cycling research has made several attempts to develop universal indices or classification systems that quantify and compare comfort levels in cycling environments. One common approach involves adapting a familiar A-F scale (Beura et al. 2021), often used in vehicular traffic analysis, to rate the perceived quality of bicycle infrastructure. Notable examples include the Cycling Comfort Index (CCI) proposed by Zhu and Zhu (2019), as well as Perceived Level of Comfort (PLOC) measurements grounded in variables such as ambient traffic, pavement markings, and traffic signs near truck loading zones (Abadi & Hurwitz, 2018). Recent contributions from Canadian studies have also sought to categorize cycling facilities based on comfort levels (Can-BICS), often incorporating regional infrastructure variations and cyclist demographics (Public Health Agency of Canada 2020; Winters et al. 2022). These models represent efforts to formalize the concept of cycling comfort into scalable, policy-friendly instruments. Despite this diversity, one metric stands out particularly dominant in literature: the Bicycle Level of Service (BLOS).

One of the most prominent and widely adopted approaches for assessing cycling comfort is the Bicycle Level of Service (BLOS) index. Originally derived from the Level of Service (LOS) framework in the Highway Capacity Manual for vehicular traffic (Morris et al. 2010), BLOS adapts this model to the cycling context by aiming to capture perceived user comfort. Although not explicitly labeled as a comfort metric in all studies, BLOS has been widely interpreted and applied as such (Kazemzadeh et al. 2020; Zhu & Zhu 2019). Several studies have traced the development of BLOS from its early origins through its refinement and broader application in urban cycling infrastructure assessments (Li et al. 2012a; Šemrov et al. 2022). A comprehensive review by Kazemzadeh et al. (2020) outlines its methodological evolution and recurring use across diverse geographic contexts. More recent versions of the index have incorporated additional factors such as slope (Furth et al. 2023), indicating an ongoing effort to adapt BLOS to real-life cycling experiences. However, some scholars have noted that BLOS may conflate safety and comfort and does not always directly reflect cyclists’ subjective experiences, particularly in complex environments like shared spaces or intersections (Blanc & Figliozzi 2016). Despite its widespread adoption, BLOS remains fundamentally grounded in objective, external factors, such as vehicle speed, lane width, and traffic volume, rather than direct measurements of cyclist comfort. As such, BLOS can be considered a universal, infrastructure-centered index that overlooks the diversity of user experiences, preferences, and physical responses. While some studies attempt to incorporate perception through surveys, they often rely on pre-defined indicators embedded within the BLOS framework, rather than allowing cyclists to define comfort in their own terms. In sum, although BLOS offers useful information about the conditions that influence comfort, it does not constitute a direct measure of comfort itself. This reflects a broader divide in the literature between studies that assess comfort indirectly, often via safety or traffic-inspired indices, and those that seek to understand comfort directly, through empirical engagement with cyclists’ own definitions and embodied experiences.

3.3.4 Explorative and methodological studies

In the walking realm, over the years, the focus of the articles shifted from a mostly objective and quantitative view to a more holistic approach. The first subjective perspectives are included from 2021 onwards (Bigazzi et al. 2021), while also the first mixed-methods approach appears (Ma et al. 2021). The only purely qualitative approach is adopted by Alsina-Pagès et al. (2024), which implements a conscious walk assessment that combines both subjective and objective views. The authors challenge people’s awareness of urban elements and their significance for the urban climate. The method of conscious walks does, therefore, not only provide a comprehensive data collection but also knowledge transfer and education (Alsina-Pagès et al. 2024).

It is noticeable that recent walking articles use new technologies such as 3D-modeling-based VR simulations in combination with questionnaires (Piao & Lee 2022), simulations with audio-visual environments in a semi-anechoic chamber (Ren et al. 2023), or Google Street View panorama picture assessments (Huang et al. 2024). Data-driven approaches are also used for the separate assessment of the environmental parameters on comfort, including thermal, visual, and acoustic aspects. While sensory parameters are often captured and analyzed through modeling and simulations (Takebayashi 2016; Teshnehdel et al. 2020), a trend goes towards on-site measurements and questionnaires, which allow recording of subjective perceptions in direct comparison to objective measurements (Liu et al. 2022a; Ma et al. 2021; Xiao et al. 2024).

A similar shift toward more explorative and methodologically innovative approaches can be observed in cycling research, though it remains limited to a handful of studies. Ayachi et al. (2015) explore this concept through an online survey of enthusiastic cyclists worldwide, directly asking what contributes to their comfort. The study identifies a combination of influencing factors: bicycle-related (frame, saddle, handlebar), environment (road type, weather), and individual (body position, bike adjustments, physical strain). In the same year, Calvey et al. (2015) use on-bike surveys, instrumented bicycles, and questionnaires to explore satisfaction, which is found to overlap with comfort and safety. It highlights that maintenance issues, especially surface conditions, are critical to users’ perceptions. A few years later, Fitch et al. (2022) apply a visual preference survey and show consistent comfort ratings across population segments, with the presence of bike infrastructure and low-speed roads emerging as key contributors. Interestingly, the study also underscores how comfort and safety are often conflated, given their strong correlation.

Two studies in our sample also offer valuable methodological insights. Fitch and Handy (2018) compared data from real-world bike rides and web-based video surveys to assess comfort and safety perceptions. The study reveals a systematic negative bias in imagined experiences: participants rated video-based rides around 10–15% less comfortable and safe than in-person experiences, suggesting that more realistic, behaviorally grounded survey methods may yield more accurate insights. The study categorizes comfort surveys as descriptive, comparative, or behavioral. In a related effort, Kazemzadeh et al. (2020) contrast cyclists’ perceptions gathered in traffic with those from online settings. They observe that comfort ratings are generally more optimistic in online surveys, but only at the operational (local) level. At broader (strategic and tactical) levels, responses remain consistent across methods. Together, these findings suggest that real-life experience tracking offers the most reliable data collection approach when studying subjective and intangible concepts like comfort.

4.3.1 Active mobility comfort framework

Our proposed framework (Figure 4) conceptualizes comfort within a socio-cultural context (macro-environment), which acts as the overarching lens for the system. This context, comprising local travel cultures, legal frameworks, social norms, and historical infrastructure standards, indirectly shapes how the immediate environment is perceived. By placing the model within the macro-environment, we acknowledge that comfort is socially situated. A physical stressor like traffic volume isn’t a universal constant; it is filtered through the traveler’s socio-cultural standing, such as gendered perceptions of space or the legal protections afforded to them in a specific driving culture. This explains why the same built environment produces radically different comfort perceptions across different demographic groups, thereby providing a theoretical basis for inclusive and equitable urban design. Within this macro-environment, three core constructs interplay: the built environment (stressors), the natural environment (regulators/modifiers) and the human processor.

Figure 4: Active mobility comfort framework: The built and natural environments (rounded rectangles) act as inputs into the human processor (shaded dashed box). Comfort (C_t) is conceptualized not as a linear output but as a state of harmony at the intersection of physical, physiological, and psychological systems (Venn diagram). Solid arrows indicate direct stimuli flow, while the dotted line represents the recursive feedback loop based on the traveler’s prior state. All processes are contained within the overarching socio-cultural context.

The human processor experiences this interplay through three synchronized layers of comfort: physical (sensory contact), physiological (biological response), and psychological (cognitive interpretation). The final momentary comfort perception (C_t) is generated dynamically over time and fed back into the system via a temporal loop, updating the baseline for the remainder of the journey.