Optimising Cyclist Road-Safety Scenarios Through Angle-of-View Analysis Using Buffer and GIS Mapping Techniques

Abstract

In the present era, achieving sustainability requires the development of planning strategies to develop a safer urban infrastructure. This study examines the realistic aspects of cyclist safety by analysing cyclists’ fields of view, using Geographic Information Systems (GIS) and spatial data analysis. The research introduces novel geoprocessing tools-based GIS techniques that mathematically simulate cyclists’ angles of view and the distances to nearby environmental features. It provides precise insights into some potential hazards and infrastructure challenges encountered while cycling. This research focuses on managing and analysing the data collected, utilising OpenStreetMap (OSM) as vector-based supporting data. It integrates cyclists’ behavioural data with the urban environmental features encountered, such as intersections, road design, and traffic controls. The analysis is categorised into Special Classes to evaluate the impacts of these aspects of the environment on cyclists’ behaviours. The current investigation highlights the importance of integrating the objective environmental elements surrounding the route with subjective perceptions and then determining the influence of these environmental elements on cyclists’ behaviours. Unlike previous studies that ignore cyclists’ visual perspectives in the context of real-world data, this work integrates objective GIS data with cyclists’ field of view-based modelling to identify high-risk areas and highlight the need for enhanced safety measures. The proposed approach equips urban planners and designers with data-informed strategies for creating safer cycling infrastructure, fostering sustainable mobility, and mitigating urban congestion.

1. Introduction

Urban transportation, as one of the fundamental avenues of contemporary research, plays a key role in responding to the needs of modern societies, especially in achieving sustainability and environmental protection goals. This field of study is firmly established in the scientific literature, focusing on improving safety and promoting clean and low-carbon transportation methods. In this regard, providing standard and safe infrastructure platforms for cyclists is of particular importance; by reviewing and modifying infrastructure design criteria, safe and efficient routes for safe cycling can be provided. Urban mobility, as one of the fundamental components of sustainable development, has a pivotal role within this field, and cycling, as one of the main pillars of active and environmentally friendly transportation systems, plays an increasing role in the realization of sustainable cities [1]. The encouragement of cycling requires the development and design of appropriate and safe infrastructure elements, a requirement that plays a crucial role in enhancing the sustainability of the transport network. Furthermore, the improvement of cycling safety can ultimately act as an essential factor in responding to the ongoing challenges facing this transport mode, thereby reducing mental and practical barriers for users, increasing usage rates, improving the quality of urban life, and supporting the broader goals of environmental sustainability. However, many previous studies have relied on general urban environmental frameworks, an approach that may overlook the subtle and context-specific challenges cyclists face, particularly in relation to perceived safety [2]. Recent studies have demonstrated that Geographic Information Systems (GIS) and well-structured spatial data provide a valid and effective approach for advanced visualisation and accurate analysis of urban environments. The present approach provides a suitable platform for identifying factors affecting cyclist behaviour in urban environments and ultimately improves safety-related metrics for cyclists in urban contexts [3]. This research gap indicates a severe deficiency in the design of urban infrastructure related to cycling routes, so key components such as safety, field of vision, and the comfort of cyclists are optimally considered. Comprehensive studies that move from a holistic to a component-based approach are needed to address these gaps. Such studies should adopt novel methods for the analysis of cyclist experience, using GIS-based buffer modelling, and real-world bicycle paths to gain insights from the cyclist’s standpoint [4]. Analysing cyclists’ perceptual experiences along urban routes provides a detailed understanding of how specific built and infrastructural elements contribute to perceived and actual safety conditions [5].

Literature Review

The use of Geographic Information Systems (GIS) to assess infrastructure planning and cyclist safety has gained increasing popularity in recent years [6,7]. GIS-based tools have proven valuable for mapping crash data, modelling preferred cycling routes, and evaluating the influence of environmental factors such as road design, intersection density, and land-use patterns on cycling behaviour [8]. These spatial analyses enable more objective evaluations of cyclist exposure to risks across different urban contexts. One of the most common methods in spatial cycling research is the application of buffer zones to simulate the surrounding built environment. Traditional studies typically employ circular or network-based buffers that are fixed around a cyclist’s trajectory or point of interest [9]. Buffer zones extract spatial features such as traffic signals, crossings, and surface types to understand the environmental influences on cyclist route choices or crash risks. However, these methods frequently assume uniform visibility and perception in all directions, which might not accurately represent what it is like for cyclists to proceed through dangerous paths. The limitations of such isotropic buffers have been brought to light by an increasing amount of research. Moreover, existing models often exclude the dynamic cognitive load involved in urban cycling, such as how cyclists visually scan for hazards, anticipate movements, or adjust cyclists’ behaviours in real time [10]. To address these limitations, recent studies have begun integrating elements of visual perception into spatial modelling, drawing on ergonomics and human-factor-based research. For instance, eye-tracking studies and helmet-mounted cameras have been employed to approximate fields of vision and attentional zones [11,12]. Even though these tools provide insightful information, there is still a lack of integration with GIS techniques. Few studies have proposed buffer models that specifically mimic an angular field of view, one in line with the orientations and speeds of cyclists. This study addresses that gap by suggesting a geospatial buffer with new shape to simulate the perceptual space of a cyclist navigating an urban setting.

Bicycles represent an environmentally friendly option: As a sustainable mode of transport, cycling offers both environmental benefits and a positive user experience. Nevertheless, safety concerns and inadequate urban infrastructure remain significant barriers to its widespread adoption in urban settings [13]. The available statistics show that cyclists are involved in many urban accidents in Europe. Accordingly, a more accurate understanding of cyclist behaviour and movement in urban environments and the development of strategies to promote safe routes are essential [14,15]. Researchers have attempted to systematically identify and model key factors affecting cyclist behaviour in urban environments using extensive field studies, data-driven analyses, and advanced simulations [16,17].

The safety and behaviours of cyclists on the road are influenced by a variety of factors, which can be broadly categorised into five key elements: (1) human factors (e.g., the behaviours of drivers, cyclists, or pedestrians); (2) traffic conditions (such as traffic density and speed); (3) infrastructure elements (including road design and traffic signage); (4) vehicle-related aspects (such as vehicle type and structural characteristics); and (5) environmental conditions [18]. Therefore, the factors influencing cyclists’ behaviours in urban environments can be identified through a dual perspective comprising human psychological processes and an understanding of environmental elements [19,20]. To improve the safety of cyclists in city environments, researchers need to consider factors such as road layout, cycling-infrastructure coverage, protection measures, and the characteristics of cyclists [21]. In the present era, research focusing on the environments around non-motorised routes has been recognised as a key component of sustainability [22]. Accordingly, the urban environment is considered a key factor in shaping citizens’ behaviour and their efforts to maintain safety; therefore, identifying and analysing environmental elements concerning specific case studies has become essential. [23,24]. Factors of the built environment (BE) can be classified into two main categories: neighbourhood-level factors (including density, diversity of use, design, and distance) [25], and street-level elements (including streetscape and environmental perception) [26,27]. A comprehensive study of built-environment variables provides more accurate and valuable data to urban designers and researchers, facilitating a more profound understanding of the correlations between physical characteristics of the environment and human behaviours, and allows the design of more optimal interventions to improve the quality and safety of urban environments [28]. Environmental factors such as safe cycle-path infrastructure are a significant barrier to cycling adoption [29].

Recent studies on the relationships between bicycle accidents and “built-environment (BE) factors” have aimed to create safer cycling environments; these initiatives can be categorised into five main areas: (1) road design consideration, (2) traffic facilities analysis, (3) traffic control analysis, (4) sociodemographic analysis, and (5) land-use analysis [30,31,32]. To design environments that better meet cyclists’ needs, research focused on cycling-specific infrastructure and urban environmental elements has expanded significantly in recent years [33]. In fact, the main goal of this study is to optimise cyclists’ road safety through the development and application of GIS-based buffers and viewpoints to provide practical simulations of how cyclists interact with their surroundings in relation to their surroundings. The results underscore the need to align infrastructure planning with cyclists’ subjective perceptions to foster safer and more accessible environments. In this context, recent studies have increasingly examined visual perception and the personal interpretation of surrounding spaces.

Urban Infrastructure and Cyclists’ Behaviours: To promote cycling, the development of safe, low-emission, and low-noise infrastructure integrated within the broader transport network plays a key role; the improvement of cycling safety can act as a fundamental factor in highlighting and responding to the ongoing challenges facing this mode of transport [34]. Indeed, such infrastructure development could lead to people being encouraged to use bicycles, the quality of life in cities being improved, and perceived barriers for cyclists being reduced, all of which support broader sustainable development goals for the future [35]. This has also attracted the attention of policymakers, leading to increased focus on implementing safer cycling routes and the improvement of walking conditions in urban spaces [36]. With respect to cycling, infrastructure policies concerning the design, location, and redistribution of spaces for active transport are shaped by cultural and political frameworks in a specific spatial context and cannot be universally generalised [37]. Considering the substantial investments in cycling infrastructure and other active transport modes [38], as well as the high costs and logistical challenges associated with redesigning urban spaces [39], it is essential that new infrastructure can attracts more users. Research indicates that subjective or perceived safety is critical in this process. In the contemporary era, allocating a proportion of safe road space commensurate with the expansion of cycling is essential; this strategy can increase citizens’ motivations for cycling by reducing mental barriers and creating security, ultimately strengthening clean and sustainable transportation and improving the quality of urban life [40]. Accordingly, while these studies contribute to the scientific discourse, there remains a need for more accurate simulations and a deeper understanding of cycling behaviour within urban environments, particularly regarding the influence of urban elements on cyclist-related dynamics [41]. The various threats and risks cyclists face in urban environments underscore the importance of examining how the built environment affects the behaviour and safety of these cyclists [42]. Extensive previous research has examined aspects of the features of the built environment, using geographic data sources [43,44]. Such observations emphasize the intersection of spatial policies and environmental psychology, highlighting subjective safety as a central concern in developing active transportation infrastructure [45,46].

Recent studies have shown that various aspects of urban infrastructure, including segregated lanes and designed intersections, directly impact cyclists’ behaviours and perceptions of safety on urban routes. These elements are key in how users evaluate and navigate urban spaces. The general connectivity of the cycling network and the connection between pedestrian zones and land-use density also affect cyclists’ choices and comfort levels. Figure 1 illustrates these common urban factors and their effects on cyclists’ behaviours in a simple framework.

Studies in this field are generally categorised into two main types: qualitative and quantitative. Qualitative studies assess cycling infrastructure, public space quality, traffic, and cycling environments. Quantitative studies evaluate the route, the surrounding cycling environment, the cyclist’s behaviour, etc. [47].

Objective built-environment attributes: Contradictory findings have been reported, depending on whether the built environment had been measured objectively or assessed based on individuals’ perceptions [48,49,50]. Objective measurements of the environment are often obtained through systematic observations based on street images or virtual audits from Geographic Information Systems (GIS) and existing spatial databases [51,52]. In contrast, perception-based assessments are usually conducted through self-report methods, such as structured interviews or standardized questionnaires [53]. However, to date, few existing studies have comprehensively examined whether and how the effects of the objective indicators and the subjective perceptions of active transportation differ. Despite the empirical importance of this issue, there is still insufficient evidence regarding the mediating mechanisms that link objective environmental measures to individual perceptions and, ultimately, to active transportation-related behaviours.

Many transportation and urban planning studies highlight the need for more reliable empirical data to understand better the constraints that influence cycling uptake [54]. Data obtained from geographic analysis tools such as spatial buffers, which are based on vector-based GPS spatial data, can effectively support hypotheses related to how cyclists perceive the environment around their routes and perform safety assessments [55]. Using Geographic Information Systems (GIS) databases also makes use of a cost-effective, practical, and accurate tool for road planning and data analysis processes. These tools can improve route safety and assist in the development of safer urban infrastructure for cyclists [56,57]. In addition, leveraging the results of the multi-criteria analysis within the Geographic Information Systems (GIS) system enables more accurate and realistic results, aiding in the visualisation of outcomes and supporting informed decision-making in the design of future road infrastructure [58]. This study demonstrated a GIS technology-based planning tool that, used with field data, can help optimise and address the potential disadvantages of cycling routes [59].

Buffer Analysis in GIS Applications: A buffer is a geoprocessing function of the Geographic Information Systems (GIS) framework that buffers spatial elements such as points, lines, and polygons. This area serves as a peripheral zone for spatial analyses such as adjacency, overlap, and built-environment assessment; it functions as a user tool in geographic analysis and urban planning [60]. Additionally, the buffer provides a platform to identify spatial features for route lines by extracting data from within the buffer [61]. From a research standpoint, buffer evaluation enhances data extraction accuracy when assessing the built environment and road networks, particularly in spatially concentrated locations [62]. Two primary buffer types are employed, namely, constant-width and variable-width buffers; these are generated through either raster-based or vector-based methods. Understanding the buffers’ visual accuracy and behaviour in Geographic Information System (GIS)-based environments is critical when simulating cyclists’ field of view. Figure 2 illustrates the differences between raster and vector representations of buffers, especially when zooming in or rendering spatial objects accurately. In the context of guaranteeing reliable and consistent spatial extraction around cyclists’ paths, this image highlights the significance of using high-precision vector buffers like those used in this study [63].

Resolution distortion may affect raster buffers, whereas vector buffers offer greater precision in representing spatial features (see Figure 2). The accuracy of buffer-based analysis is significantly improved when factors such as orientation, visibility, and variable buffer dimensions are considered—particularly in complex urban environments [66]. Furthermore, the buffer generated in a GIS environment with a geoprocessing tool allows the user to optimize the data structure of the layer on which it is applied. The results of the buffer analysis provide an optimal hybrid geoprocessing framework for the analysis of spatial data at different scales. Experiments using real datasets show that considering buffers with respect to their radius and orientation is essential for accurate investigation of the built environment, as these factors not only increase the accuracy of spatial analyses within the environment and associated maps but also enable the identification of complex behavioural patterns and prediction of high-risk spots, and ultimately provide evidence-based recommendations for improving the safety and sustainability of urban infrastructure [67]. Advanced buffer modelling incorporates mathematical algorithms and computational techniques to handle large datasets and consider real-world conditions, such as road obstacles or a cyclist visibility constraint [68]. This approach is popular due to its data-centric focus and simplicity of implementation [69]. Previous researchers have sought to analyse cyclists’ behaviours, including psychological factors and how people perceive the environment based on the number of accidents and environmental factors that affect cyclist safety; however, many of these studies have neglected accurate simulations and user-perception-based data due to their focus on statistical data or simple modelling [70,71]. Moreover, researchers believe that a significant focus on those objective criteria can also ignore people’s subjective impressions motivated by cognitive processes [72]. Studies reveal that cycling accidents are influenced by intersection location, building dedicated cycling lanes, and street hierarchy [65]. Thus, a thorough investigation is required to identify the impacts of these attributes and their connections with a person’s attitude or influencing factors. To assess the cycling environment, evaluation indicators have been developed based on two primary data extraction resources: (1) a comprehensive review of the literature on environmental elements affecting cycling routes, and (2) the use of open-source data to collect both quantitative and qualitative information relevant to infrastructure and detailed spatial assessments [73]. In sum, the reliance of earlier approaches on methods that have excluded large-scale empirical data and simulation-based methods often leads to design shortcomings and inadequate urban analysis [74]. Previous approaches generated buffers without accounting for the cyclist’s field of view and relied on overly simplified geometric forms. In contrast, the present study uses a systematic analytical framework to simulate the behaviours of cyclists in urban environments. It examines the impacts of physical features and components of the urban environment on the affected safety-related and behavioural patterns, on a neighbourhood scale. This research contributes to a more comprehensive analysis of the relationship between environmental factors and cycling system dynamics by providing practical insights. However, the approaches used in this study have limitations because they focus mainly on a limited set of variables, such as certain environmental features or specific cyclist behaviours along the route, and do not include a comprehensive and integrated analysis of the surrounding space. On the other hand, studies utilizing a holistic approach and determining the elements that cyclists perceive as being on the path, based on the visual and spatial characteristics perceived by cyclists, can provide a comprehensive understanding of the path and the obstacles, and other environmental factors in the urban environment that directly affect cyclists. This study proposes a novel buffer configuration intended to simulate a cyclist’s forward field of view, defined by buffer with new shape. The aim is to extract and analyse environmental features within this perceptual zone and to evaluate cyclist safety.

2. Materials and Methods

Objective features of the built environment have been used to identify and categorise environmental factors related to the built environment; comprehensive studies must adopt more realistic approaches [75]. Past studies have comprehensively addressed the processing of roads and built environments based on environmental variable analysis, aiming to evaluate the objective environmental data using Geographic Information Systems (GIS) technology [76]. Furthermore, spatially referenced data can be obtained through extraction and processing in a GIS environment and used as a valuable resource for accurate spatial analysis and evidence-based decision-making [77]. Previous research utilizing GIS-based spatial analysis has identified major cycling routes and infrastructure gaps and enabled the extraction and processing of empirical data through GIS [78]. In fact, effective use of bicycle data requires GIS data processing to perform comprehensive analyses such as QGIS, aiming to extract comprehensive analyses for built-environment processing and the determination of accurate spatial data [79]. OpenStreetMap (OSM) provides spatial data on urban infrastructure and road networks, which can be extracted and visualised as vector layers in QGIS for realistic mapping and spatial analysis [80]. These studies address the perception and measurement of environmental features [81]. In urban environments, objective measurements are typically derived from quantitative assessments. The use of geographic information system (GIS)-based tools allows for more accurate assessments of safety indicators, as well as deeper analyses of environmental features, both visual and non-visual, by providing a platform for a systematic and comparative analysis of physical and spatial features, which can lead to the generation of practical insights for optimizing the design of urban infrastructure and improving the safety of active transportation users [82]. Accordingly, a relationship between the perceptions and the objective measurements of the same environmental features, and the impact of this relationship on factors relevant to active transportation, has been established. Perception-based moderation mechanisms have also been investigated [83]. However, the previous research has not adequately captured the effects of authentic user experiences or incorporated empirical field data from cyclists [84]. From a broader perspective, the previous research in this field has rarely examined the influence of the built environment on user behaviour using empirical evidence, and the scarcity of such studies is recognized as a significant gap in the literature [85]. This research presents a conceptual framework that examines the urban environment’s effects on cyclists’ behaviours. Based on past research studies, it is hypothesised that environmental factors along cycling routes can influence cyclists’ behaviours. The analysis is structured around key categories of built-environment elements, including streets, intersections, buildings, and other urban features. To extract the “big data”, a tool such as GIS is needed to filter and analyse the data related to cycling routes relevant to features that exert a significant impact on cyclists’ behaviours. Based on empirical spatial modelling and the analyses conducted in QGIS, this method examines how the built environment influences cyclist behaviour.

This analysis involved classifying urban infrastructure components into the vector data types of points, lines, and polygons, filtered based on their relevance to cyclist safety and behavioural outcomes. The classification model, illustrated in Figure 3, was developed to estimate the influence of environmental features on cyclists’ behaviours by distinguishing between primary variables (directly affecting behaviour) and contextual variables (indirect environmental influences). The underlying assumption is that the perception of urban environmental elements is directly influenced by cyclists’ behaviours.

These environmental components were categorised into distinct classes, each representing a subtype within the urban context, as defined by having similar spatial characteristics. These include route typologies, points of interest, and other quantifiable urban metrics.

This approach leverages open-source mapping tools and buffer-based spatial filtering to identify two critical classification layers:

Primary Class: A model based on cyclists’ viewing angles and distances, used to assess behavioural influence.

Special Class: Descriptions of environmental features within the cyclists’ perceptual fields along the route.

This class-based typology supports scalable modelling across diverse urban contexts and facilitates the extraction of behaviourally significant spatial data.

Outlining the effects of elements of the urban environment on cyclists’ behaviours: First, as previously mentioned, it is assumed that cyclists’ perceptions of various urban environmental elements influence their behaviours. These typologies are further assumed to represent Special Classes that reflect different subtypes of urban environments, each defined by characteristic spatial and functional attributes. Such elements can be extracted from route data, mapped points of interest, and relevant urban metrics. Moreover, more advanced models can also be developed by identifying and analysing these subgroups by utilizing multiple data extraction and classification methods.

The methodology for analysing urban environments and the elements along cyclists’ routes employs buffer tools to estimate two key classification layers:

(i)

Primary class: A model for assessing cyclists’ behaviours, based on forward viewing angle and distance.

(ii)

Special class: Descriptions of environmental features located within the cyclist’s perceptual field along the route.

This classification process is supported by open-source mapping tools and buffer-based spatial filtering, aiming to identify the primary criteria and relevant variables across various urban contexts (see Figure 3).

Modelling a cyclist’s field of view and considering the positional and directional parameters provides a realistic representation of spatial perception and helps to improve route design and enhance infrastructure safety. Accordingly, the present study, focusing on cyclist-related data, introduces this category as one of the principal axes of analysis and, by examining behavioural data and users’ subjective perceptions, provides a basis for designing evidence-based interventions in sustainable transportation.

The Primary Class refers to cyclist-centred spatial criteria, specifically the “field of view and distance to surrounding features.” This class models cyclists’ spatial perceptions by using parameters such as yaw angle, eye-level visibility, and angle–distance buffer zones. It is designed to estimate how cyclists observe and respond to their surroundings while navigating urban routes.

The Special Class “encompasses the environmental elements” located within a cyclist’s perceptual field, as defined by the Primary Class. This includes built environmental features (e.g., intersections, buildings, vegetation, and road elements) derived from OpenStreetMap and extracted using QGIS buffer geoprocessing analysis.

So, this study classifies analysis into two core components: Primary Class and Special Class. The Primary Class focuses on cyclist-centred variables such as field of view, distance, and yaw angle to model how cyclists perceive their immediate surroundings. The Special Class includes environmental features (e.g., roads, buildings, and traffic signs) that fall within this visual field, extracted using QGIS buffers analysis from OpenStreetMap (OSM) vector data. These classes integrate perceptual data with objective spatial elements, enabling a realistic assessment of how urban environments influence cyclists’ behaviours.

• Primary Class

The proposed model based on variables evaluating cyclist behaviour, and the assessment features associated with cyclists’ behaviours, have been limited to two criteria. The cyclist’s position, geographic location, and directional heading relative to a fixed reference point determine a cyclist’s visibility along a route.

The model incorporates the following variables:

• Variables

LATO, LONO: The latitude and longitude of the observer or reference point.

LATB, LONB: The latitude and longitude of the bicycle (cyclist).

YAW: The heading/yaw of the observer.

D: The visibility range/distance (e.g., 25 m).

• Formula

The visibility condition is determined by calculating the cyclist’s relative bearing from the reference point and assessing whether this bearing falls within the defined visibility angle. Based on the two primary parameters, distance and field-of-view angle, these criteria can be categorised into distinct classes, forming the foundation for two preliminary modelling approaches. The first, and more efficient, model is developed based on the analysis of cyclist behaviour and models the distance and viewing angles in response to environmental changes in the urban context. In addition, the model defines a visual function that can be used to assess the quality of the spatial and directional parameters affecting cyclists’ spatial perceptions and situational awareness. As shown in Table 1, an advanced visual-field analysis model is proposed, one designed based on spatial and directional variables; this model aims to assess the degree of visual accessibility along cycling routes. This model includes two complementary analytical approaches:

-

The first method concentrates on how cyclists behave and considers changes in the field of vision and distance in response to shifting urban environmental factors. Any environmental feature along a bike path has the potential to obstruct or restrict the visible area, which could impact safety and decision-making.

-

The second approach assigns visibility values to spatial parameters along the path, allowing for systematic assessments of the impacts of environmental features on cyclists’ visual perceptions.

Table 1. Improved model describing visibility relative to the cyclist.

Criteria	Description
Cyclist’s Eye Level and Environmental Obstacles (D_obstacle)	Calculate the sight distance to the horizon and consider any intermediate obstacles.
Dynamic Angle of Visibility (Θ_dynamic)	This angle can dynamically change based on the environment. For example, in a city, buildings might limit the angle significantly.
Visibility Function (V (Θ, D))	This is a function that incorporates both angle and distance to determine the quality of visibility.

Table 1. Improved model describing visibility relative to the cyclist.

Criteria	Description
Cyclist’s Eye Level and Environmental Obstacles (D_obstacle)	Calculate the sight distance to the horizon and consider any intermediate obstacles.
Dynamic Angle of Visibility (Θ_dynamic)	This angle can dynamically change based on the environment. For example, in a city, buildings might limit the angle significantly.
Visibility Function (V (Θ, D))	This is a function that incorporates both angle and distance to determine the quality of visibility.

• Specific-Class

Many studies have been published showing significant bicycle-use growth in recent years. These studies have examined the characteristics of the built environment using data from sensors and advanced analytical tools and have adopted a performance-based approach to the analysis and evaluation of urban cycling-related infrastructure [86]. Therefore, examining the data related to the built environment as constituting a special class in the research can emphasize the necessity of paying attention to and integrating these data with behavioural analyses and pave the way for evidence-based urban interventions [87]. In previous studies, the primary focus has been examining cycling routes, an approach similar to previous research that requires considering a variable such as the angle of deviation for more accurate calculations. [88]. The present research will identify elements in urban environments that influence cyclists’ behaviours. In this regard, the field-of-view radius and yaw angle (the angle between the observer’s line of sight and the horizontal plane) can be computed from the yaw angle’s computation and the change in the bicycle’s direction.

• Variables:

r: Radius of the visibility circle (maximum distance the observer can see).

YAW: Observer’s yaw (heading) angle, in degrees.

θ: Angular width of the visibility sector, centred on YAW in the radius buffer. Typically, this is 60°, as described earlier.

• Example of a Visibility Sector:

The observer’s yaw (heading) is YAW= 90° (which corresponds to east).

The cyclist has a visibility range r = 100 m.

The cyclist has an angular visibility width of θ = 10°.

This approach provides a realistic understanding of route assessment by analysing cyclists’ viewing angles and perceptual distances while identifying environmental parameters affecting behaviour and safety. This process paves the way for infrastructure design that is more accurate, and the optimization of safety interventions based on empirical data.

The analytical framework is divided into two main segments: First, the expected values for the main category are calculated. Then, the environmental parameters for the area surrounding the cyclist’s path are evaluated. Furthermore, the research criteria are classified into two dimensions: The first dimension is human centred, and focuses on cyclists’ perceptions, reactions, and behaviours in the face of the urban environment. And, in an environment-centred dimension, the framework analyses the features of the built environment of the area along the route, using spatial data in point, line, and polygon formats. As outlined in

Table 2. An algorithm and a detailed description of the pseudocode formula for this method are provided.

Step	Algorithm Expectation—Maximization to Estimation UEACB
1	Initialise (LON, LAT, DISTANCE, ANGLE, Yaw)
2	E-C Step: Estimation of Criteria
3	Using geographic coordinates, which allows a bicycle proprietor to be properly positioned, is an effective technique for direction guidance.
4	Using yaw records for navigation is essential in identifying the bicycle behaviour direction of motion.
5	Geometry and a particular component help determine the ideal distance by which the attitude is decided.
6	End: The angle ratio can be computed based on a complete 360-degree view, such that if the angle (ANGLE) represents the cyclist’s clear line of sight; it can precisely indicate the angle inside which the cyclist can observe clearly.

, the algorithmic structure and corresponding pseudocode for this two-stage analysis illustrate how the simulation integrates cyclist perception with environmental constraints. This procedural framework estimates which spatial elements will become perceptible and behaviourally significant at various points along the route.

Suppose a cyclist has an eye level of 1.6 m above ground and the road ahead has a slight downward slope, giving an angle of depression of α = 5°:

$$\begin{gathered} \mathrm{D}={\displaystyle \frac{1.6}{\mathrm{Tan}(5^{\circ })}}\approx 18.3\mathrm{m} \\ \mathrm{Angle} \mathrm{Ratio}={\displaystyle \frac{120}{360}}={\displaystyle \frac{1}{3}} \end{gathered}$$

• Spatial Data

The data section involves collecting, preparing, and processing the data related to cyclists and urban infrastructure, which will be managed in a GIS environment. This study uses tools like OpenStreetMap (OSM) and QGIS software (version 3.28.14) to collect, organise and analyse the data. This section describes methods for refining and processing records related to the urban environment, including environmental elements and routes travelled by cyclists. The data collection and analysis process employed in this study is structured into three main phases:

I.

Data Collection (This phase consists of three components):

-

Component A: A comprehensive review and collection of large-scale spatial data from OpenStreetMap (OSM), forming the foundational dataset. As illustrated in Figure 4, the number of vector features associated with each stage is specified.

-

Component B1: Environmental features surrounding the cycling routes are extracted using the QuickOSM plugin in QGIS, leveraging publicly available open-source spatial data.

-

Component B2: Additional environmental data are extracted, specifically, from within the buffer zone that represents the cyclist’s field of view, aligned with the observed trajectory.

II.

Data Preparation (this phase involves two key operations):

-

A. Data Integration: Merging multiple datasets to construct a unified geospatial framework.

-

B. Data Filtering: Selecting and refining features based on predefined criteria relevant to bicycle routes and their immediate surroundings.

III.

Data Processing and Simulation (in the final phase):

-

The actual cycling route is overlaid with simulation data, using QGIS.

-

A perceptual buffer is created along the trajectory to simulate cyclists’ field of view and enable high-resolution modelling of real-world conditions.

-

Spatial features located within this buffer are extracted and analysed to identify behaviourally significant urban elements.

-

Cyclist behavioural data are integrated with built-environment characteristics through Volunteered Geographic Information (VGI) and GIS-based mapping methods.

Figure 4. The processing of the data.

Figure 4. The processing of the data.

Buffer analysis, as a fundamental technique in geographic analysis, is instrumental in detecting the elements of the built environment that cyclists interact with along their routes.

Design and Application of a Simulation Buffer: This study designed a buffer with an angular geometry (60 degrees, range 25 m) based on the cyclist’s absolute position, aiming to simulate cyclists’ perceptual field of view along the route realistically. This design aims to extract environmental elements within the field of view and analyse their impacts on cyclists’ behaviours and safety. This approach provides a suitable platform for accurately and systematically assessing cyclists’ interactions with the urban environment.

Simulation Buffer (a buffer creation based on the cyclist’s field of view): To extract data using the buffer calculation formula based on the application of viewing angle and actual field of view, it is assumed that it is possible to extract the elements of the built environment that influence the behaviours and decisions of users along the path. In this approach, this buffer was constructed and analysed within QGIS, where it was applied to model cyclist–environment interactions. The simulation considered both the distance between the user and surrounding features and the directional field of view, thereby enabling realistic spatial extraction of the built-environment elements that are behaviourally relevant to the cyclist.

As illustrated in Figure 5, the buffer geometry model was centred on the cyclists’ absolute position and was used to simulate the cyclist’s field of view and analyse corresponding environmental and behavioural data. The blue line shows the direction of travel, and the purple dashed line shows the cyclist’s central orientation. As shown in Figure 5, the geometric model of a buffer with an angle of 60 degrees and a radius of 25 m, located at the cyclist’s absolute position, represents cyclists’ usual field of view. The red sector specifies the spatial extent of the field of view, and the red line shows the radius of the angle, which determines the length of the buffer. The main goal is to evaluate the effects of environmental elements in this 60-degree field of view on cyclists’ behaviours and safety consequences.

The 60-degree angle, based on the previous formula, provides a realistic representation of the cyclist’s field of view and is consistent with the principles of visual perception in dynamic urban contexts.

To demonstrate the buffer tool’s potential, simulations were conducted within an urban neighbourhood to replicate the spatial conditions experienced by cyclists. First, data collected by users is merged using the QGIS tool, and then it is overlapped with the actual route in the city, providing comprehensive coverage of the neighbourhood and route. Following the method described, the geographic coordinates of these points are used to extract the actual data from OSM and collect data about the urban environment at each location.

The radius and angle chosen for the buffer are designed to ensure the necessary stability to accurately capture important environmental functions around the study point while limiting computational requirements to a predictable and manageable level. For many applications, particularly those in urban or roadway environments, the selected radius enables the collection of nearby environmental data without the inclusion of excessive or irrelevant elements. The field of view defined in this study is designed based on a common human visual focus in urban spaces and effectively simulates what cyclists or pedestrians might perceive directly in their surrounding environments. By considering perceptual boundaries in spatial analysis, this approach realistically reflects the human experience along cycling routes. Accordingly, the performance of the angular and distance-based buffer has been evaluated based on human perception criteria and a typical sight radius ranging within 25 m of the route. Furthermore, angular buffers show a high level of consistency in extracting the environmental data related to the sight range of cyclists on urban routes. By combining directional perspective and perceptible spatial radius, this method enables an objective analysis of users’ environmental exposure based on distance and field of view.

To further support the modelling framework, empirical research on visual perception, cycling reaction time, and urban complexity led to the selection of a 25 m radius and a 60-degree angular buffer. A cyclist’s perceptual range and cognitive processing needs during navigation are realistically approximated by this configuration, which was not chosen randomly.

The model captures spatial proximity, and the mental workload associated with real-time decision-making in dynamic road settings, by aligning the buffer dimensions with the ways in which cyclists perceive and react to their surroundings. Regarding data sourcing, OpenStreetMap (OSM) was used as the principal environmental dataset. While OSM is a community-driven database and may show variability in its coverage, the dataset was filtered and validated through a multi-step process.

This included the following steps:

(1)

Querying only tagged and geocoded features with high confidence ratings.

(2)

Using satellite orthophotos and satellite image base maps to verify key elements along the selected routes.

(3)

Excluding outdated or irrelevant geometries through attribute-based and geometry-based filtering in QGIS.

The extracted spatial features were categorized by geometry type (point, line, and polygon) and organized according to their potential impacts on cyclist safety.

Table 3. Summary of vector geometry types and their relevance in cyclist safety analysis.

Geometry Type	Feature Type	Data Source	Relevance to Cyclist Behaviour
Point	Traffic lights, benches	OpenStreetMap (OSM)	Influence stopping behaviour and attention distribution
Line	Road edges, cycling paths	OSM/QGIS	Guide trajectory, affect speed and comfort
Polygon	Buildings, vegetation zones	OSM/Satellite Imagery	Impact visibility, perceived safety, and navigational choices

, below, summarises this categorization, indicating the types of features, the data sources, and the relationships of these features to behavioural responses when within the cyclist’s perceptual field of view.

This methodology facilitates the spatial extraction of environmental features within a cyclist’s forward field of view, offering a structured and replicable framework for assessing how infrastructural variables affect cycling behaviour and safety.

2.1. Data Collection

Data collection constitutes a fundamental component of Geographic Information System (GIS)-based spatial analysis frameworks [89]. Spatial tools facilitate this process by enabling descriptive representation of available data for planners and highlighting areas requiring further investigation and monitoring [90].

As illustrated in Figure 6, the data collection framework is organized into three main stages:

• Acquisition of large-scale spatial data from OpenStreetMap (OSM), which includes streets, buildings, and other urban elements relevant to cyclist safety.
• Targeted extraction and querying of environment-specific features using the QuickOSM plugin in QGIS, enabling focused filtering for areas along designated cycling routes.
• Integration of cyclist-generated geospatial data, such as GNSS (Global Navigation Satellite System) tracks, with spatial layers used to develop a comprehensive geodatabase for further analysis.

Figure 6. Data collection steps in the research.

Figure 6. Data collection steps in the research.

Collecting and refining large-scale datasets makes it possible to identify the critical infrastructure and behavioural factors required to model cyclists’ behaviours effectively. This process enables a more accurate assessment of the primary determinants influencing how cyclists interact with their environment.

2.1.1. Data Collection: OpenStreetMap

This section examines the use of open-source mapping tools to analyse urban spaces and dedicated cycling routes within the study area in a city. The necessary datasets for this research were extracted through the refinement of available data from open-source platforms and the application of spatial simulation tools [91,92]. Specifically, OpenStreetMap (OSM) and QGIS were employed to collect and classify spatial data related to the urban environment. The OSM data were categorised into three vector types, namely, point, line, and polygon; these represent various features, such as built-environment elements and cycling infrastructure.

As a source of volunteered geographic information (VGI), OSM offers a reliable depiction of real-world urban infrastructure and road networks.

Data elements represent physical features of the urban environment, such as roads, trees, buildings, and traffic lights, along with their corresponding geographic locations. These features are typically classified into multiple categories (e.g., major roadways, lampposts, and fences) and may include detailed attributes such as road classification, number of lanes, or traffic signal specifications. This section examines the use of open-source mapping tools to analyse urban spaces and dedicated cycling routes in the selected case study. The data required for the research will be extracted by refining the available data from open-source resources and using simulation tools. One of the open and accessible sources of volunteered geographic information (VGI) samples is the OpenStreetMap (OSM) project, which is provided by users with different levels of mapping experience and made possible by the availability of economic mapping tools and easy-to-use software [93]. Several studies have determined that urban characteristics such as land-use mix, the greenway width, building density, and parks and public squares are significantly associated with cycling behaviour [94]. Areas exhibiting a diverse land-use mix are more likely to encourage cycling by accommodating a range of urban functions and travel purposes [95]. A key objective of cyclist-oriented risk management is to analyse the local urban environment and design routes that align with cyclists’ safety needs, preferences, and behavioural patterns [96]. Unlike previous studies, this research integrates a structured classification of environmental features from OpenStreetMap with cyclist-specific data obtained through shape-based buffer analysis, providing nuanced insight into the environmental factors influencing cycling behaviour along urban routes.

2.1.2. Data Collection: Quick OSM

This study selected an urban area as the focal point for data extraction using the Quick OSM plugin within QGIS. This selected area was simulated in order to facilitate spatial analysis, and the relevant data collected within this boundary was subjected to further evaluation.

A sample of OpenStreetMap (OSM) data was downloaded on 2 April 2024, using a predefined set of OSM parameters. The dataset covers a large area of approximately 2.98 square kilometres (2,896,643.792 square metres). Spatial information concerning cyclist round-trip routes within this designated area was extracted. QGIS was employed as the primary tool for processing and managing geospatial data. Initially, 632 distinct OSM elements were extracted for each location. To ensure analytical relevance and reduce dimensionality, the OSM data was filtered based on the urban functionality of each feature.

2.1.3. Data Collection Utilizing a Buffer

Buffers, used as tools in data mining and spatial analysis, enable targeted data extraction and the analysis of the spatial relationships between environmental elements by determining a certain radius and distance [97]. Using buffer analysis in transportation networks enables a more precise depiction of the actual spatial conditions and facilitates in-depth assessments of the interactions between infrastructure and users [98]. Buffer technology includes raster- and vector-based approaches, each offering specific analytical capabilities and enabling a wide range of spatial analyses [99]. Previous methods utilised buffer techniques for data mapping and extraction; with the creation of a circle with a border on the map, the buffer can help urban planners to extract the data and elements defined within this border separately and thereby understand the buffer zone [100]. Using the buffer tool in QGIS, it is possible to create defined areas around key spatial features that can be intersected with related layers through geographic processing tools. Given that the focus of this study was the impact of the environment around cycling routes, all spatial information within a 25 m radius of the recorded locations was extracted using the buffer defined in the proposed framework [101,102]. Additionally, in the review of the previous research and studies, it was found that most have primarily focused on data extraction using circular or polygonal buffers. This highlights the need to develop a more accurate simulation approach that incorporates a novel buffer configuration based on geometric formulas that account for the cyclist’s viewing angle and distance.

2.2. Data Preparation

Data preparation is a fundamental step towards achieving accurate and realistic modelling and simulation of cycling routes and analysing cyclists’ interactions with the environment. This phase involves integrating and filtering large datasets obtained from open-source geographic databases, GPS recordings, and GIS-based tools [103]. Using QGIS, cartographic data are managed as georeferenced vector layers, facilitating the calculation and analysis of these quantitative environmental indicators [104,105].

As part of the proposed methodology, one approach involves identifying environmental elements; these are categorized into two primary groups:

-

The first group includes types of infrastructure such as roads, cycling lanes, and other built-environment features, and is obtained from OpenStreetMap (OSM).

-

The second group includes specific attributes of cycling infrastructure, such as the distinctions between car roads and dedicated cycle paths, road surface quality, and relevant natural or environmental features. In this stage, the analysis addresses the question of how such data are to be prepared and categorized to support the analytical framework of this study (see Figure 7).

2.2.1. Data Merging

Merging the various datasets in this research was one of the most critical steps, not just technically, but also conceptually. The analysis utilized a combination of data sources, including GPS cycling tracks, GNSS data files, and environmental datasets from OpenStreetMap, each stored in a distinct file format: GPX, Excel, Shapefile, or OSM. Once the data was fully integrated, it became possible to examine how environmental variables influence cyclists’ behaviours. This integration in QGIS enabled spatial analyses that closely reflected real-world conditions within urban networks. For instance, isolating environmental elements within a cyclist’s field of view, and positioned inside designated buffer zones, provided a novel and evidence-based approach for designing safer cycling infrastructure.

The data integration process was carried out in the QGIS software (version 3.28.14) environment. As can be seen in Figure 8, this process began by importing the GNSS recorded route data and GPX files into the QGIS environment and then aligning these routes with OpenStreetMap (OSM) layers, allowing for a precise overlap between the actual cyclists’ routes and the environmental elements. This alignment enabled the targeted and analytical extraction of built-environment elements along the routes. It is worth noting that the OSM database included features related to cycling infrastructure, such as parking facilities, repair stations, and service areas, which were integrated into a single spatial dataset for further analysis [106,107].

• Use OSM and QGIS to integrate data related to cyclist behavior analysis and infrastructure data. This merging process requires multiple file formats, including GPX, Excel, Shapefile, and OSM (OpenStreetMap XML format):
  • First, cycling path data integration begins by importing GPS cycling path data in GPX format.
  • Second, GNSS data for the cycling path is imported from Excel format.
  • Third, environmental OpenStreetMap data is loaded in Shapefile format.
  • Fourth, using GNSS data and GPX files in QGIS initially poses a challenge because these datasets do not align correctly due to shape discrepancies. The ‘Overlap’ command in QGIS transforms datasets into an accurate rent cycling route. This integrated route data can subsequently be used for detailed analysis.
• The cycling path analysis in this study involves extracting data from open sources such as OpenStreetMap and merging it within the QGIS environment. A significant challenge is the initial discrepancy between GNSS sensor data collected from cyclists and the actual bicycle route, resulting from inaccuracies in satellite sensor connections.
• Explanation of Subfigures (Figure 8):
• Subfigure (a) illustrates the raw GNSS data (red dashed lines), which does not align precisely with the cycling path.
• Subfigure (b) demonstrates the overlay process, showing how GNSS data (red dashed line) and GPX Track data (blue dashed line) are compared and initially aligned, highlighting the discrepancies.
• Subfigure (c) displays the final merged result, where GNSS data and GPX tracks have been successfully integrated to accurately represent the true cycling path (green dashed line). This processed data can be a reliable basis for further spatial and behavioral analysis.

Once the datasets were fully integrated, targeted spatial analyses could be conducted. For instance, environmental elements within the cyclist’s visual field or designated buffer zones were identified to assess potential hazards and inform safer routing strategies.

2.2.2. Data Filtering

Filtering data based on the mapped features can affect the accuracy of findings regarding the built environment and elements close to the cycling route. Removing duplicative data, or data that is not related to the routes used by cyclists and lacks spatial validity, completes the data-cleaning process. This provides a final and reliable set for analysis. This step allows for the accurate extraction of key features of the built environment that affect cycling behaviour [108]. It is worth noting that environmental elements, especially urban buildings, can negatively affect the accuracy of GPS point registration with respect to cycling routes [108]. Using spatial filters effectively separates and isolates data related to the specific infrastructure of the cycling route [109]. The data cleaning process prevents the re-registration of duplicate elements inside the path boundary. It paves the way for a more accurate identification of the environmental elements that affect cyclists’ safety, perceptions, and decision-making [110].

Several essential technological steps were taken during the data preparation process to guarantee methodological validity and spatial precision. All GNSS tracks from the cyclist’s path were first overlaid onto OpenStreetMap (OSM) base layers in QGIS to verify alignment with known urban landmarks and infrastructure. Coordinate snapping and georeferencing tools were used to resolve differences in terms of the absolute positioning between the OSM features and the gathered GPS data. Redundancy and possible spatial noise were eliminated by removing duplicate entries and overlapping geometries. All GPX, CSV, or SHP format data sources were reprojected to a standard coordinate-reference system (WGS 84/EPSG:4326 as geographic coordinates and EPSG:3857 as projected cartographic coordinates) to ensure consistency across spatial layers. Additionally, environmental features extracted using the QuickOSM plugin were filtered and cross-validated using orthophotos and satellite-image base maps to ensure their correspondence with real-world locations. A thorough cleaning and integration process was conducted to enhance the accuracy of the spatial overlays used in the field-of-view simulations and buffer modelling and the reliability of the extracted features in interpreting interactions between cyclists and their environments.

Filtering spatial data based on relevant map features is essential to improving the precision of findings related to cyclists’ interactions with the built environment. As illustrated in Figure 9, the web-based OpenStreetMap (OSM) platform is the principal source for extracting urban environmental data within the study area. To ensure the accuracy and precision of the analyses, raw data must be systematically filtered to remove irrelevant, repetitive features, as well as those features that are not directly related to the cycling route.

Data filtering serves two primary objectives:

(1)

To eliminate redundant or non-representative features that could compromise the accuracy of behavioural analyses, such as duplicate GPS points or infrastructure unrelated to the cyclist’s trajectory.

(2)

To isolate and retain only those spatial elements, such as buildings, road typologies, urban greenery, and transportation infrastructure, that directly influence cyclists’ perceptions and safety.

As illustrated in Figure 10, QGIS-based tools enable the spatial refinement of urban environment layers, allowing researchers to isolate features within defined areas and angles relevant to cyclist visibility. This methodology facilitates a precise evaluation of the infrastructure elements that cyclists experience within urban environments.

This approach to data filtering comprises three principal stages:

• Initial Feature Identification: Environmental data is initially extracted from the OpenStreetMap Wiki through a key-value schema.

This step involves organizing spatial data into major categories, including transportation infrastructure, land-use classifications, amenities, and natural features, thus providing an essential dataset for subsequent analytical procedures.

2.

Feature Classification and Refinement: The extracted spatial data are classified and refined utilizing QGIS software in conjunction with the QuickOSM plugin, enabling precise categorization of relevant geographic features.

Features most relevant to cyclists’ behaviours, such as road types, building types, urban vegetation, and bicycle infrastructure, are retained and reclassified under context-specific labels (e.g., ‘Sidewalk’, ‘Barrier’, and ‘Amenity’).

3.

Directional Buffer Filtering: A novel buffer form is applied in QGIS, shaped by the cyclist’s angle of view and route proximity. This buffer is used to isolate the features visible and behaviourally relevant to cyclists, ensuring high fidelity in behavioural analysis.

Environmental data were first extracted from the OpenStreetMap database through key-value pair queries to identify and systematically categorize features related to the urban environment. Elements influencing cyclists’ experience and behaviour, including road networks, buildings, and green spaces, were refined and filtered using the QuickOSM tool in QGIS. Next, a dedicated buffer based on the cyclist’s viewpoint was applied to separate spatial data within the actual field of view. Applying a systematic approach to environmental data refinement allows us to focus on the elements most relevant to cyclists’ perceptions, safety, and mobility-related behaviours. A greater understanding of how urban ambient elements impact cycling behaviour is made possible by incorporating the cyclist’s actual field of view into the spatial analysis, which improves realism. Ultimately, this approach supports the creation of safer, more bike-friendly urban areas by offering insightful information to legislators and urban planners.

2.3. Data Processing

Data processing within the current research approach involves a structured, step-by-step methodology, whereby the accuracy of result evaluation depends significantly on the careful processing and preparation of index data. The initial step in this sequence consists of filtering the OpenStreetMap (OSM) data, using QGIS software.

2.3.1. The Loading of the Data Map for the Location and the Simulated Path of the Bicycle

Simulation is essential to understand how urban built environments influence cyclists’ behaviours and to support the design of safer cycling routes within cities.

This study employed two complementary modelling classes: an environmental model that defines spatial contexts and identifies specific urban elements, and a cyclist behaviour model that simulates cyclists’ dynamic responses based on viewing angles and distances to the environmental features encountered along their routes (see Figure 11).

Each cyclist followed a predefined round-trip route, equipped with advanced data acquisition systems, including Global Navigation Satellite System (GNSS) receivers and GIS-integrated sensors, facilitating real-time spatial data collection. The surrounding built environment was categorised and visualised within QGIS according to spatial typologies; the spatial data used included linear data (such as road networks and boundaries), point data (such as traffic signs, traffic lights, barriers, and facilities), and polygon data (such as buildings and green spaces). Each geometric type was assigned a distinct colour code for accurate visual interpretation and to make spatial analysis more effective.

The processed data were then used as input in a classification-based simulation, with the primary class representing cyclist behaviour and the secondary class representing environmental elements affecting the perceptions and safety associated with the route.

In the proposed scenario, each cyclist rides a pre-determined round-trip route, using bicycles equipped with advanced data collection systems including Global Navigation Satellite System (GNSS) receivers and sensors integrated with a Geographic Information System (GIS). This configuration allows for real-time spatial data recording and provides the conditions necessary for detailed analysis of the route environment. The built environment along the cycling routes is analysed through distinct spatial typologies, with lines, points, and polygons extracted from OpenStreetMap (OSM) and processed using QGIS software.

Specifically, the linear data represents the network of streets and street boundaries; the point data includes discrete elements such as barriers, traffic signs, traffic lights, and urban facilities; and the polygon data includes larger areas such as buildings and green spaces. Each spatial category is assigned a distinct colour code to facilitate accurate visual identification and the efficient processing and interpretation of the spatial data.

2.3.2. The Real Path of the Bicycle

An additional environmental simulation is needed to make the route realistic. Other data from the findings can be integrated by adding the information to the QGIS software. Data such as GNSS collected by the cyclist (using a bicycle equipped with various tools for data collection) can be added to the real-path stage. In fact, simulating the cyclist’s route with the QGIS tool, merging the data with the real route in the urban environment, and, finally, ensuring the data’s accuracy by checking the ID and YAW angle and longitude and latitude angles can achieve cleanness and correctness with respect to the data. After determining the cyclist’s path and collecting relevant data, the extracted data is processed through a directional buffer to identify environmental elements that may influence the cyclist’s behaviour. By representing the cyclist’s actual field of view while moving along the path, this buffer allows the extraction of spatial features that fall within cyclists field of vision. In this manner, visual elements of the built environment potentially influencing the cyclist’s perception and behavioural responses are systematically separated and analysed.

This research operationalized the combination of subjective cyclist perception and objective GIS data by establishing directional buffer zones that mimic the cyclist’s forward field of vision. These angular buffers, which were set at 60 degrees with a 25 m viewing radius, were designed to represent the spatial boundaries of a cyclist’s perceptual field when navigating an urban setting. Based on empirical research on perceptual attention during cycling, environmental elements within these simulated cones were assumed to fall within the cyclist’s visual awareness. Thus, instead of collecting direct survey data, the model incorporated perception indirectly through geospatial proxies, allowing objective environmental data (e.g., intersections, barriers, and traffic signals) to be analysed within a realistic cognitive context. This perceptual–spatial overlay was the foundation for the interpretation of cyclist behaviour and risk exposure along the route.

This study’s 60-degree field of view is based on established research on human frontal vision and cyclist situational awareness. This angle approximates the typical horizontal field of active visual attention for individuals in motion, particularly cyclists navigating complex urban settings. Using GIS-based tools, the parameterized buffer was constructed as a directional cone extending 25 m from the cyclist’s yaw orientation, capturing the spatial extent within which cyclists are most likely to detect and respond to environmental stimuli. Using angular buffers provides a more accurate simulation of cyclists’ perceptual exposure than traditional circular buffers, as this approach is consistent with the real-world conditions of safety-related decision-making, which occurs in the context of spatial perception and directional vision. This approach also allows for the extraction of environmental data concerning the cyclist’s field of view in complex urban situations.

2.4. Example Application of the Proposed Buffer

To demonstrate the functional implementation of the proposed 60° buffer, a segment of the cyclist’s route was selected from the simulated area and is shown in Figure 12. The buffer identified key infrastructural elements within this visual field, including three intersections, two pedestrian crossings, and several roadside barriers. These components were spatially concentrated close to high-density crossings and at turning points, suggesting that cyclists have higher behavioural and perceptual demands. On the other hand, buffer segments applied to straight paths with dedicated cycle lanes recorded fewer environmental features, indicating a lower cognitive load. This comparison highlights the buffer’s effectiveness in isolating the infrastructure components most likely to affect cyclist decision-making and safety.

To increase the realism of the cycling route representation, a supplementary stage of environmental simulation was performed. In this stage, additional spatial data, specifically, GNSS tracks collected by cyclists using sensor-equipped bicycles, were integrated into the QGIS environment to accurately refine the alignment of the simulated route with real urban infrastructure.

Integrating spatial data, including absolute coordinates (latitude and longitude), track ID (ID), and yaw angle, is essential to ensure spatial accuracy and validate the results of the simulation.

After determining the cyclist’s path, a directional buffer with a 60-degree angle and a range of 25 m was designed to simulate cyclists’ perceptual field of view and provide a realistic analysis of the interaction with the surrounding environment (Figure 12).

This method allows for the removal of irrelevant parts of the built environment outside the perceptual range while at the same time providing a more accurate assessment of the elements influencing cyclists’ perceptions and decision-making.

The buffer allows for selective querying of spatial features—such as traffic signs, urban greenery, building fades, or obstacles—within the visible path. These elements are recorded as vector data (point, line, or polygon features) and filtered spatially within the buffer zone. Their effects on how cyclists act may change depending on how close they are to intersections or important city landmarks. This shows the importance of looking at risk and perception in context. To simulate cyclist perception along the route, the yaw angle representing the rider’s direction of movement was derived from sequential GPS coordinates, using directional vector analysis in QGIS and GIS technology. The angle was calculated by analysing the changes in position between successive GPS points to determine the cyclist’s heading at each moment.

A sample segment of the cycling route, displayed in Figure 12, was examined to provide a practical illustration of the proposed buffer’s utility. The directional buffer identified a cluster of infrastructural elements, such as the two intersections, three pedestrian crossings, and one roadside barrier within proximity along a curving section of the route. These elements are concentrated where cyclists must make rapid decisions, suggesting a higher likelihood of behavioural adaptation or risk. In contrast, buffer zones applied to long, uninterrupted straight paths with dedicated cycling lanes captured fewer environmental stimuli, indicating a relatively lower perceptual and behavioural load. This contrast supports the hypothesis that the proposed angular buffer effectively isolates areas of elevated infrastructure-related complexity. While a comparative analysis with conventional circular buffers is beyond the scope of this study, future research should explore such validation to reinforce the applied value of this model.

For instance, simulated bicycle trajectories could be analysed using directional buffers based on angle and distance, enabling the extraction of spatial data that approximate the visual perspectives of cyclists in varied urban contexts.

The various elements that appear as vector data inside the buffer zone with a specified radius (angle and distance) directly affect cyclists’ behaviours. Also, depending on the context of the built environment, these elements have a greater or lesser probability of affecting cyclists’ behaviours based on the proximity of these elements to intersections or landmarks.

The yaw angle, which represents the directional orientation of the cyclist’s line of sight along the route, was extracted from the real-time GPS trajectory data using GIS-based tools in (QGIS). Each GPS coordinate point was matched with a corresponding heading angle derived from changes in the spatial coordinates over sequential timestamps. This yaw angle thus represents the forward-facing direction of the cyclist at each segment of the route.

The field of view (FOV) was defined based on a conical buffer with a central yaw orientation, a 60-degree angle span, and a 25 m radial distance. These parameters were chosen based on validated human field-of-view standards found in the literature for moving cyclists under urban conditions, considering peripheral and central vision ranges. The field of view was simulated using directional buffers in QGIS and calibrated to represent the realistic visual envelope of a cyclist while in motion.

Regarding empirical validation, although this study did not include live user testing (e.g., wearable device eye-tracking experiments), the modelling framework was informed by empirical values cited in previous road-safety and human perception studies. This includes experimental research on cyclist behaviour in response to environmental stimuli and works in standard observational fields from the ergonomic and transportation psychology literature [36,65].

With the possibility of simulating other types of sustainable transportation in various combinations of influential urban environment elements, the applicability of the method and the utility of the framework incorporating buffering by angle and distance are underscored. On one side, the framework supports the analysis of urban settings and the features that distinguish safe routes from high-risk cycling routes. From one standpoint, this framework enables the identification of urban environments and the characteristics that define safe or risky routes for cyclists. This process is enhanced by an advanced buffer that includes the angle of view and the optimal distances to surrounding environmental elements along the route.

This issue, often overlooked in previous research, is addressed here for the first time through a novel buffer approach based on the cyclist’s field of vision and perceived distance.

To further illustrate the application of directional barriers,

Table 4. One example of a directional buffer (60° angle, 25 m radius) is seen in the capture of a real-world intersection and obstacle (e.g., building or barrier) within the cyclist’s field of view. This illustration highlights spatial components relevant to perception and behavioural response.

Obstacle Type	OSM Tag/Data Source	Geometry Type	Behavioural Relevance	Example Impact
Roadside Barrier	barrier	Line	High—Influences perception of safety	Narrowing lanes or limited manoeuvrability
Parked Vehicles	parking =, amenity = parking	Polygon/Point	Medium—Visual occlusion, perceived collision risk	Obstructs sightlines at intersections
Walls	building	Polygon	High—Major visibility obstruction near intersections	Obstructs line-of-sight, increasing risk during turns or when merging
Vegetation (Trees/Hedges)	natural = tree, barrier = hedge	Point/Line	Medium—Aesthetic or visual obstruction	Reduces visibility, especially on curved paths
Signposts/Bollards	highway = street_lamp, bollard	Point	Low to Medium—May obstruct narrow paths	Perceived as physical risk in constrained spaces
Construction Area	landuse = construction	Polygon	High—Safety risk and deviation from planned route	Unexpected obstacles requiring rerouting
Garbage Bins/Benches	amenity = bench, amenity = waste_basket	Point	Low—Spatial clutter in constrained cycling lanes	Forces detours or speed adjustments
Raised Curbs/Medians	highway = traffic_calming, kerb	Line	Medium—Affects turning angles and balance	Forces cautious manoeuvring
Fence/Railings	barrier = fence, barrier = rail	Line	Medium—Restricts lateral movement	Limits emergency avoidance space
Unpaved Surface Patch	surface = gravel, surface = unpaved	Line/Polygon	High—Affects stability and decision-making	Cyclists may slow down or change route

presents an example of an intersection scenario in which a visible obstacle, such as a physical barrier, is located in the perceptual field of a cyclist. This visualization illustrates the identification and modelling of key elements of the urban environment near the perceptual barrier and provides empirical evidence for analysing the behavioural impacts and assessing associated risks.

To extend the methodological innovation proposed in this study, a specific application of directional buffer analysis is presented for the modelling of cyclist perception in complex urban contexts. The model uses a 60° angular field and 25 m radius to simulate a realistic perceptual cone of a cyclist, allowing for the accurate extraction of visible urban elements along the route. By integrating orientation and perceived distance, this approach goes beyond traditional circular buffers, representing the actual field of view of a cyclist navigating urban environments.

As seen in the intersection scenario, numerous physical obstacles such as walls, fences, parked vehicles, and vegetation can be located within the angular buffer zone and significantly affect cyclists’ fields of view, movement behaviours, and decision-making processes at critical intersections.

The integration of spatial typologies—point, line and polygon features are essential in simulating cyclist perception from a behavioural and infrastructure perspective. As shown in Figure 13, spatial data within the buffer were analysed based on cyclists’ functional roles. For instance, line data (e.g., highway = cycleway) represent the structure and connectivity of the route; these are critical for predicting movement behaviour and safety levels. polygon data such as the designation of green spaces (landuse = grass) reflect environmental quality, while point data such as pedestrian crossings (highway = crossing) and traffic lights directly influence cyclists’ decisions and behaviour along the route.

The dual classification model is divided into a Primary Class (perceptual criteria such as viewing angle, distance, and deviation) and a Special Class (urban elements within this perceptual range). How can real-world environmental obstacles, after passing through this angle-distance buffer, be directly linked to changes in perceived risk? The answer to this question strengthens the claim that perception-based spatial modelling, which is based on field-of-view analysis, enables a richer understanding of behavioural drivers in complex and multifaceted urban networks.

Ultimately, the results presented in Figure 13 closely align with the visual data extraction steps described in the Materials and Methods and Results sections. Visual verification of buffer-based extraction, applied to point, line, and polygon data, supports the model’s accuracy and validates the cyclist-centric GIS framework. As the study highlights, this method makes it possible to simulate urban perception under realistic constraints. In order to create safer and more welcoming cycling environments, it highlights potential risks, infrastructure flaws, and spatial irregularities that need to be fixed.

The proposed approach constitutes a data mining framework grounded in the perception of cyclists’ behaviours along urban routes. It is structured in two stages: first, environmental and behavioural elements are classified based on cyclist perception and objective spatial data (classification stage); second, directional and distance-based parameters are computed using GIS-based buffer modelling techniques. After this, the data is evaluated in a spatial analysis framework to assess consistency and accuracy.

3. Results

According to the studies and reviews conducted, the buffer modelling method, which emphasizes spatial relationships and directional parameters for cyclists, can effectively simulate the urban environment, including both the areas surrounding the routes and the routes themselves, by utilizing data collected from OpenStreetMap, QGIS, and GIS.

In this study, an advanced GIS-based analytical framework was implemented using QGIS software and OSM data to enhance the perception of the cycling route and systematically identify environmental elements influencing the cyclist experience along the path.

By generating directional buffers based on angle and distance parameters that simulate the cyclist’s visual perspective, relevant urban environment features associated with cyclist safety and behaviour were extracted and analysed.

The extraction process resulted in discrete mapped features, including the following:

-

Linear features (e.g., highway, pedestrian, bicycle, right-of-way, left-of-way, service, surface, barrier, separated paths, and sidewalk).

-

Point features (e.g., amenities, natural elements, traffic lights, highway points, and crosswalks).

-

Polygon features (e.g., buildings and land-use areas).

The dataset covers a large urban area and contains an extensive amount of vector data. The data was filtered based on relevance and proximity to the bicycle path. Key features identified within the buffer zone shaped by the cyclist’s field of view were ultimately extracted, including physical barriers, intersections, dedicated bicycle lanes, and adjacent land-use categories.

The preliminary analysis in the present study shows that fixed urban features (such as buildings and fixed obstacles) in a cyclist’s field of view in urban areas affect cyclists’ behaviours along the route. Two categories, Primary and Special Class, were applied to analyse cyclists’ behaviours, and the characteristics of the built environment, within the buffer framework, highlighting how specific environmental elements influence cyclist decision-making.

Building upon the established methodological framework, the classification into Primary Class and Special Class provides the structural foundation for the spatial analysis conducted in this study. To simulate how cyclists perceive cyclists’ surroundings while in motion, the Primary Class defines cyclist-oriented perceptual parameters such as field of view, yaw angle, and vision range. In parallel, the Special Class refers to the spatially extracted urban features such as intersections, road segments, sidewalks, buildings, and land-use areas that fall within the cyclist’s calculated visual buffer.

This dual classification, using spatial data mining based on the integration of perceptual modelling, allows for a more accurate assessment of how elements of the built environment affect cycling behaviour in the real world.

Data collection, filtering, and buffer estimation using innovative techniques are recognised as valuable in identifying which components of the urban environment have the most significant influence on cyclists’ behaviours. This study extracted vector-based spatial elements from the simulated buffer zones within QGIS to identify the key urban infrastructure components affecting cyclists’ behaviours.

As detailed in

Table 5. Final map data extractions within the buffer, with angle and distance.

Feature Name	Geometry Type	OSM Tag	Relevance to Cyclist Safety
Highway	Line	highway	High—Core infrastructure for routing
2. Foot	Line	foot	Medium—Pedestrian-priority path
3. Bicycle	Line	bicycle	High—Cycling infrastructure
4. Sidewalk Right	Line	sidewalk = right	Medium—Affects lateral space perception
5. Sidewalk Left	Line	sidewalk = left	Medium—Affects lateral space perception
6. Sidewalk	Line	sidewalk	Medium—General pedestrian infrastructure
7. Service	Line	service	Medium—Minor roadways, influence conflict
8. Surface	Line	surface	Medium—Surface condition affects safety
9. Barrier	Line	barrier	High—Separates traffic, improves safety
10. Segregated	Line	segregated	High—Indicates dedicated cycling infrastructure
11. Footway	Line	footway	Medium—May be shared with cyclists
12. Amenity	Point	amenity	Medium—Supportive infrastructure
13. Natural	Point	natural	Low—Aesthetic or minor spatial influence
14. Traffic Signal	Point	highway = traffic_signals	High—Controls movement
15. Highway (Signal/Point)	Point	highway	High—Signalised control points
16. Crossing	Point	highway = crossing	High—Major decision and risk zones
17. Building	Polygon	building	Medium—Affects enclosure and sightlines
18. Land Use	Polygon	landuse	Low—Background environmental context

, the extracted features were grouped by geometry type, namely, line, point, and polygon, and represent elements that can influence cyclist perception, decision-making, and safety.

To address the identified research gap, this study undertook a focused literature review emphasizing the development of a novel buffer geometry and an innovative methodology for simulating and perceiving the urban environment from the cyclist’s experiential perspective. Recent studies on enhancing cyclist safety have increasingly considered the interaction between human factors and environmental variables [111]. Building upon this foundation, the present study integrates open-source spatial data with cyclist-generated inputs to comprehensively analyse the ways in which built environmental features and contextual elements influence cyclist behaviour [112].

The previous research has demonstrated that geoprocessing techniques, particularly buffering along the routes travelled by cyclists, can yield valuable real-world insights into the urban environment [113]. However, modelers have used geometric buffering methods to assess urban environments but have generally not considered the cyclist’s experiential viewpoint [114]. The lines features include pathways such as highways, bicycle lanes, and sidewalks, as well as infrastructure attributes like surface type and physical segregation. Point features capture discrete objects like traffic signals, pedestrian crossings, and natural or built amenities. Polygon features denote larger spatial zones, including buildings and land-use areas. These categories were derived from OpenStreetMap and validated within the QGIS environment after merging with the GNSS data collected by cyclists.

This classification is the basis for our simulation. It lets the system know which spatial parts are visible to the cyclist and how they might affect behaviour and safety. Angle- and distance-based buffers help us to understand spatial influence in a more nuanced manner. They go beyond traditional geometric buffers by using perceptual modelling based on the cyclist’s experience.

Subsequently, the urban environmental elements extracted through buffer geoprocessing analysis can be used to assess route conditions as experienced by cyclists, identifying key factors that influence the behaviours of cyclists. Using a GIS-based approach with OpenStreetMap (OSM) data and integrating the data into QGIS software, this study was able to extract built-environment data by applying viewing angles and visual distances that identify elements easily, and in a manner similar to human vision, from the front axis of the bicycle path (see Table 5), with vector features fully included. The study area covers the bicycle path, and includes spatial data filtered according to the cyclists’ paths and as related to safety and user behaviour. This analysis is divided into two key categories: the main factors influencing cyclist behaviour (distance and viewing angle) and specific environmental elements near cycling routes, including road design and land use. Therefore, considering the study’s objectives, which were to acquire realistic data on cyclist safety on a given route, as well as the context of responding to the global goals of encouraging cycling as an active and sustainable transportation option for a healthy environment, and noting that increased safety measures can lead to greater acceptance of cycling and creating healthier communities, these goals can be achieved using real-world, GIS-based data, and its application in the fields of urban sustainability, helping to create a global standard for road safety.

4. Discussion

Previous studies on the functions of humans, traffic, infrastructure, and elements of environmental aspects have helped to enhance cycling safety. This work investigates how built environmental elements affect cyclists’ behaviours and safety by combining visual perception modelling with spatial analysis. The suggested approach shows the spatial heterogeneity of the city. It offers a more accurate knowledge of the elements influencing cycling risks, using a field-of-view simulation and buffer analysis in a GIS environment. This approach creates a link between perceptual and spatial data, and the aim is to offer evidence-based insights and a context-based, safer cycling-infrastructure design. Paying attention to what cyclists see and experience helps identify which elements of the surroundings truly matter. Recognizing these influences allows planners to reduce risks and make cycling safer. In the long run, this approach supports bike routes that are more inviting, encouraging people to ride more often and drive less, and thereby helps cities to develop a sustainability-focused lifestyle.

The current investigation utilised existing data as independent variables representing urban environment factors (such as infrastructure elements, barriers, and traffic signs) that influence cyclists’ behaviours. These factors also include cyclist viewing angles and distances from environmental obstacles, which were investigated using GIS analysis techniques and buffering. The dependent variable measures cyclists’ behaviours, encompassing safety on bicycle paths and route choice, which are influenced by infrastructure characteristics and environmental perceptions.

This paper follows previous research that has used urban elements as key factors in mental interventions. This paper provides evidence of the effect of urban environmental elements on cyclists’ behaviours. The effects of urban environmental elements on cyclists’ behaviours have been investigated as variables, using elements and criteria (environments) based on environmental characteristics and cyclist observations.

At the outset of the process, this method categorises urban environmental typologies and systematically documents cyclists’ inherent variations. Unlike conventional models that directly measure the impacts of individual features, this approach generates separate environmental classes depending on perceptual and spatial properties. Then, within every class, the combinations of elements of cyclists’ behaviours and perceptual data give a complex understanding of risk factors and safety results.

Environmental analysis mines the relationships between urban design, cyclists’ behaviours, and safety, using GIS technology-based methods and field-of-view simulations to assess how environmental factors influence cyclist decision-making. Combining several methods and using geoprocessing buffer analysis in the GIS software (version 3.36), with spatial data from OpenStreetMap (OSM), this research highlights how cyclists interact with cyclists’ environment, focusing on determining which factors influence route choice, how they do so, and the overall safety of cyclist users. The effects of urban environmental elements on cyclists’ behaviours are described within two classes and sub-models: class-primary and class-specific.

While traditional circular buffer methods assume a uniform distribution of the influence field in all directions and thus ignore the directional nature of the cyclist’s visual perception, the 60-degree parameterized fan buffer introduced in this study provides a more accurate representation of the cyclist’s behavioural and perceptual patterns by aligning with the natural human field of view while cycling. This angular approach enables spatial filtering that reflects only those environmental elements likely to be perceived during forward motion, thereby reducing data noise from irrelevant features behind or beside the cyclist. Unlike circular buffers that assume that environmental interactions are the same in all directions, this approach uses perspective-based simulation to provide a more accurate and relevant picture of how cyclists perceive and interact with the environment from a human perspective. This plays an important role in improving the quality of urban safety analyses for cyclists, considering a possible real field of vision experienced by the cyclist, depending on cyclists’ direction of travel.

To ensure conceptual consistency, this discussion continues the earlier framework by referring to the “Primary Class” as the set of spatial perception parameters, including angle of view and vision distance, used to simulate the cyclist’s field of view. The “Special Class” corresponds to the environmental features captured within that view, such as roads, buildings, and intersections. These two classes are used throughout to organize and analyse the interactions between cyclist perception and built-environment elements.

In class-primary, cyclists’ visibility and distances to elements were analysed to assess the variables associated with cyclists’ behaviours, while in class-specific, elements in urban environments were examined. In the Special Class, the data collection process began with a large dataset, which was subsequently filtered based on proximity and the relative significance of environmental elements along the cycling route. In the following stage, our buffer framework combined a spatial classification of urban environments with perceptual data from cyclist field-of-view simulations. Using a dual-component system enables both the extraction and classification of built environmental features through QGIS and OpenStreetMap data and the quantification of the spatial influence of these features on cyclists’ behaviours and the perceived risk.

Through this framework, a two-step analysis is conducted: first, the typologies of urban areas associated with high-risk cycling scenarios are identified; second, the degrees of influence of proximity, angle of view, and visibility constraints on cyclists’ behaviours and interactions with infrastructure are assessed.

A 60-degree angular buffer provides a perceptually grounded representation of the cyclist’s forward field of view. This method increases the behavioural relevance of the extracted features by aligning spatial analysis with the natural limits of human vision during movement. Elements within this constrained visual cone, such as barriers, intersections, or traffic signals, are more likely to influence real-time behaviour, such as braking, lane positioning, or route deviation. Therefore, incorporating the cyclist’s field of view into Geographic Information System (GIS) analyses strengthens the link between spatial modelling and behavioural analyses and provides more practical insights for urban planning based on safety improvement.

In actuality, the novel field-of-view buffer with spatial extent is employed to demonstrate how such knowledge can be utilized to predict and analyse cyclists’ behaviours in both urban and off-road environments. This example directly applies the specific buffer model to examine and hypothesize the extent to which cyclists’ behaviours influence the identification of environmental elements that significantly affect these cyclists’ behaviours and safety. In this context, vector data associated with urban environments can be examined, and the characteristics of the route travelled can be simulated to help visualise, predict, and, ultimately, understand the impact of environmental elements on cyclists’ behaviours. Although the risk of accidents and risks along the path of cyclists in places with no reported cycling accidents is minimal, such realistic mapping and simulation tools serve as an approach that can help map out actions and locate places with higher potential for safer cycling.

Cyclist safety depends not just on traffic or road width but also on the kind of infrastructure cyclists ride through and the barriers they encounter. Riding across intersections, sharing lanes with cars, or navigating spaces without marked bike paths can increase risk and uncertainty. On top of that, obstacles like parked cars, fences, bollards, or even trees can block a cyclist’s view or make it harder to steer safely. This study highlighted such features when defining the cyclist’s field of view and recognizing what cyclists see while riding. This approach reveals how cyclists perceive the road environment and where potential risks may call for design adjustments.

A realistic urban environment with accurate spatial data is essential for applying classification schemes and directional buffer techniques based on angle and distance. To achieve this, simulations were conducted within a selected urban area characterized by infrastructure elements such as cycling lanes and road markings. Environmental data were extracted from open-source platforms, including OpenStreetMap (OSM), and the route was visualized in QGIS, using geospatial lines derived from GNSS tracking. This spatial representation enables the identification and localization of built-environment features along the cyclist’s path. Such an approach demonstrates the applicability of this method for analysing urban environments in the context of cyclist perception and safety.

Using a buffer model with a specific shape will allow us to simulate the cyclist’s field of view and viewing angles with respect to the road conditions more realistically. Identifying 18 feature types describing the elements along the cyclist’s path emphasizes cyclists’ importance in shaping cyclists’ behaviours and safety. The findings particularly support previous studies highlighting the roles of intersections and urban environment density as factors influencing cyclists’ behaviours. Past research efforts have employed geometric barriers without considering the cyclist’s line of sight, but this study’s 60-degree viewing angle provides a more accurate assessment of the environment’s impact. This solution solves a methodological flaw in the literature and provides cycling-focused urban planning tools. Although the model does not isolate the individual effect of each urban feature on cyclist safety, it captures the collective impact of the built environment’s complexity in the contexts of decision-making and risk perception. Thus, planners can use this method to map current vulnerabilities and simulate how infrastructure modifications may influence cyclist-related outcomes. The results indicate an entire spatial arrangement of the environment relevant to cyclist safety, hence underlining the need for integrated buffer simulations in active urban design. The model captures the overall impact of environmental complexity on decision-making and risk perception, even though it does not separately describe each individual influence of every urban feature with respect to cyclist safety.

While this study successfully identified and spatially extracted 18 features of the built environment relevant to cyclist safety, the analysis did not include a ranking or weighting of these elements based on cyclists’ relative influence. The current framework emphasizes spatial perception and scenario simulation rather than predictive modelling. However, future research could employ techniques such as the Analytic Hierarchy Process (AHP), multi-criteria decision analysis (MCDA), or supervised machine learning models to assign relative importance or weights to each feature. Such methodologies allow for a more quantitative assessment of environmental risk variables and increase the model’s capacity to prioritize infrastructure improvements based on bike safety outcomes. Although this study focuses on built-environment elements such as infrastructure and safe routes, it does not consider dynamic variables like speed or temporal congestion patterns. These dynamic conditions significantly affect cyclists’ behaviours and safety. These factors were disregarded to keep the emphasis on simulating the actual spatial view from the cyclist’s viewpoint. Future work should integrate bicycle-based data to provide a more comprehensive assessment of the crash risks associated with environmental elements relative to cycling specifically and perform quantitative calculations.

One major limitation of the current analysis is the assumption of flat terrain when modelling the field of view of cyclists. When riding in real-world conditions, changes in slope and elevation can significantly affect visibility. For instance, because of vertical obstructions and the cyclist’s forward-leaning posture, uphill sections may reduce the visible range, while downhill gradients may increase it.

While elevation data were not incorporated into the present analysis, future research could substantially enhance the accuracy of field-of-view simulations by integrating topographic profiles derived from digital elevation models (DEMs) that are compatible in terms of nominal cartographic scale with the real urban conditions of the cycle paths.

Integrating such data will provide a more accurate understanding of cyclist behaviour in diverse urban contexts, allowing for the creation of buffers to adjust visibility and direction to suit the descriptions of different angles.

To summarize, the extraction of these elements of the built environment, which are ultimately described from the perspective of the cyclist, can be used in this holistic classification (vector data) relative to key elements of the urban environment such as highways, sidewalks, cycle paths, footpaths, crossings, and traffic lights; all contribute significantly to the formation of movement patterns. Infrastructure features such as surface type, barriers and segregated paths affect accessibility and safety, especially for pedestrians and cyclists. In addition, surrounding buildings, land use, natural features, and amenities affect how cyclists behave along the route.

Finally, the present approach shows how such knowledge can be used to predict and analyse cycling safety across an entire neighbourhood or urban area. Further, this case study demonstrates how this model can be used to hypothesize how the built environment exerts an effect relevant to cyclist safety.

5. Conclusions

This study examines methods used to enhance cyclists’ safety and analyses cyclists’ fields of view using Geographic Information Systems (GIS) and spatial data analytics. This approach, as a matter of fact, distinguishes itself from the previous research using advanced geoprocessing functions applied with topographic accuracy to replicate cyclists’ visual angles and the distances from environmental features along cyclists’ paths.

The research indicates the correlations between urban design, cycling behaviour, and safety through GIS-based methodologies and field-of-view simulations using a buffer. The new buffer model is an essential approach for urban planners to utilize in ascertaining the environmental factors that affect bike behaviour and formulating ways to enhance safety. To improve cycling safety, this study focuses on the influence of infrastructure on traffic dynamics and the analysis of environmental elements. Using such an approach to identify safe or hazardous elements can provide future research with critical insights from real datasets and human perceptions of cycling routes to improve safety and identify built-environment elements that affect users’ behaviours along the route.

Although the present model provides insights into cyclist visibility and static environmental conditions, the inclusion of dynamic data such as volume and speed in future studies could significantly enhance the accuracy and applicability of safety assessments, given the demonstrated primary purpose of the data, which is based on the optimization of road safety for cyclists through perspective-based analysis using GIS technology.

Considering the main objective, clearly demonstrated in the data, which is based on optimizing road safety for cyclists through viewpoint analysis using GIS technology and extracting built-environment data with a buffer, significant results can be achieved. By using spatial data extraction from OpenStreetMap, spatial data (such as intersections, roads, or barriers) that exist in the built environment can be integrated in the form of geospatial datasets describing the real built environment. Also, along the path, the cyclist observes these elements, and so this data can be acquired by a bicycle equipped with data collection tools and GIS technology in a real urban environment and based on real testing by a cyclist to simulate the real route and collect data.

Finally, the QGIS software was employed to analyse the route and integrate the environmental data extracted from the built environment with the cyclist-related user data. In this study, the environmental features were obtained by applying directional buffers that reflect realistic viewing angles and perceptual distances, grounded in the experiential perspective of cyclists who encountered and perceived these elements along cyclists’ route. These elements can affect users’ behaviours when they change cyclists’ routes or continue along cyclists’ paths; as well, the perceptions of these components significantly impact the safety of dedicated bicycle roads.

This study’s methodology relies on two integrated analytical layers: the “Primary Class”, which models cyclists’ visibility and perception-based parameters (such as angle and distance of view), and the “Special Class”, which includes the built-environment features observed within that visual field. These classes provide the foundation for our buffer-based analysis and are consistently applied throughout the research.

The main findings and contributions of this approach, titled “Optimising Cyclist Road-Safety Scenarios Through Angle-of-View Analysis Using Buffer and GIS Mapping Techniques”, are as follows:

• The effects of urban features on cyclists’ behaviours were identified, showing how factors like intersections, barriers, and road elements influence cyclists’ safety.
• Effectiveness of angle–distance buffering: Incorporating angle and distance in buffer modelling improved the realism of the analysis and enabled the correct simulation of cyclist–environment interactions.
• Precise Spatial Representation of the Built Environment: Using exact spatial data (vector data: lines, points, and polygons), the study presented comprehensive knowledge of the built-environment components, thereby enabling an evaluation of their impacts on cyclists.
• Practical Simulation for urban planning: The method successfully simulated cycling routes to assist in the better planning of routes; this technique effectively simulated cycling paths, providing increased safety and efficiency.

This evidence confirms that urban designers can find high-risk areas and suggest evidence-based enhancements of cycling infrastructure by combining cyclist perceptual modelling with spatial data analysis. The findings demonstrate the effects of urban design and environmental factors on bicycle safety and behaviour. These studies indicate that comprehending the factors of the built environment that influence cycling behaviour and decision-making might enhance urban routes and infrastructure, encouraging active transportation. This paper introduces a valuable and applicable approach for urban designers and planners, aiming to simulate cyclist behaviour on a real roadway, identify high-risk areas, and ultimately provide evidence-based insights for the sustainable design of safer and more efficient cycling infrastructure.

While the current study focused on spatial extraction and simulation of cyclist visibility relative to built-environment features, it did not include inferential statistical analysis to assess the significance of the observed differences. Methods such as t-tests or ANOVA could provide additional validation to determine whether environmental elements, such as intersections or barriers, have statistically significant effects on cyclist safety outcomes. Future studies integrating cyclist incident data or perception-based survey responses could apply such statistical approaches to strengthen the quantitative evaluation of risk factors.

The proposed methodology centres on the use of advanced geoprocessing functions in a GIS environment and on a 60-degree angular buffer and 25 m radius; this model is designed to be replicable across urban settings, using standard GIS-based tools. Using open-source datasets (e.g., OpenStreetMap) and accessible geoprocessing techniques ensures that the framework can be adapted for similar studies in other cities. However, this study has limitations, including the lack of dynamic variables such as real-time traffic flow and elevation data, and the absence of statistical validation of environmental influences, in addition to a relational data structure external to the GIS software to better manage the information. These factors should be addressed in future research to enhance the robustness and applicability of the model.

Future work: Future analyses will be extended to other cities where comparable data is readily accessible. This will allow us to determine whether similar area environmental data is discovered in various cities. Second, environmental data in other regions can have the same effect, assisting cyclists. Finally, new, increasingly optimal results can be explored using different QGIS.