Analysis of walkability surrounding the future bus rapid transit lines in the city of Coimbra

Abstract

This study evaluates the walkability around the bus rapid transit (BRT) system, Metrobus, which will commence operations in Coimbra, Portugal, in 2025. As part of the first phase of an electric mobility system, Metrobus will traverse the city along two primary axes, reshaping urban mobility patterns. Given this shift, assessing the pedestrian environment around BRT stations is crucial to anticipate potential increases in walking trips. This research employs the OS-WALK-EU tool, which operationalises walkability through four sub-indices – amenity proximity, pedestrian-network permeability, accessible green/blue space and residential density – while incorporating a slope-dependent reduction of the effective pedestrian catchment. Using open data sources, the study demonstrates that OS-WALK-EU is an effective method for assessing walkability, with potential applications in other urban contexts. The findings reveal significant asymmetries in walkability along the BRT corridors, highlighting areas where infrastructure improvements are needed to enhance pedestrian accessibility. These insights are particularly relevant for city planners and policymakers, as they underscore the importance of targeted urban interventions to maximise the benefits of the new mobility paradigm. Ensuring high-quality pedestrian access to Metrobus stations will be essential for leveraging sustainable mobility and fostering a more walkable and connected urban environment in Coimbra.

Keywords

walkability assessment bus rapid transit OS-WALK-EU sustainable urban mobility pedestrian accessibility open data analysis

1. Introduction

Walking is the most fundamental form of mobility, shaped by cultural, personal and environmental factors (Mehta, 2008). It is not only a mode of transport but a prerequisite for social inclusion and public concerns. Yet urban barriers can undermine this fundamental right (Presidência do Conselho de Ministros, 2023). As a response, walkability, which assesses how easily urban areas support pedestrian access to essential needs (Barros, 2014), is key to sustainable urban design. Examining measurable spatial features and subjective experiences (Ariffin et al., 2021; Blečić et al., 2020) ultimately defines liveable spaces designed with people in mind (Szibbo et al., 2012).

Historically, public spaces have long mirrored societal change, shaping our everyday lives. Since the 1920s, rapid urban expansion has forced cities to address complex issues in employment, supply and transportation, while the rise of private car travel has disrupted the once harmonious shared use of streets (Amaral, 1969; Sousa, 2013). These historical dynamics explain why walking, once the dominant mode of mobility, has lost ground to motorised transport.

The legacy of this transition continues to shape cities today. The prioritisation of car-oriented planning has contributed to urban sprawl, reduced vitality and reinforced sedentary lifestyles (Mouratidis, 2021). As a result, urban mobility and public health face new pressures, many linked to the increasing level of sedentarism and car dependency (Chakrabarti and Shin, 2017). Because mobility quality directly affects urban life, sustainable mobility strategies prioritising pedestrians are increasingly central to planning agendas (Distefano and Leonardi, 2023). This shift fosters a more conscious and people-centred interaction with urban space, leading to interventions that better serve public interests (Ros-McDonnell et al., 2024).

The study of walkability plays a crucial role in this transformation, offering significant benefits for both the population and the urban environment, and walkability assessments must be integrated into different phases of urban planning, ensuring that pedestrian-friendly policies and infrastructure deliver effective and meaningful results (Abastante et al., 2020). Sustainable urban mobility is a multidimensional challenge that requires simultaneous attention to its economic, social and environmental aspects (Ngossaha et al., 2024). Mobility, at its core, is the ability to move freely and when applied to an urban setting, it defines how people and goods are transported (Costa et al., 2017). This capacity is fundamental for individual and collective well-being and promoting urban quality of life and social inclusion (Elsamani and Kajikawa, 2024). Compact design has been associated with higher walkability and reduced greenhouse (Lo and Houston, 2018; Talen and Jeong, 2019).

The literature identifies several core dimensions of walkability:

Feasibility – the physical capacity to walk, influenced by factors such as age, physical condition and the presence of children (Alfonzo, 2005).

Accessibility – the spatial distribution and connectivity of land uses and the physical and psychological barriers that may hinder mobility (Blečić et al., 2020).

Safety – the perceived security within the urban environment, shaped by proper lighting, diligent maintenance and active public spaces. Notably, the lack of gender-sensitive planning can undermine women's rights to safe public spaces (El-Kholei and Yassein, 2022).

Comfort – the ease and pleasure of walking, which is enhanced by suitable infrastructure and effective traffic-calming measures (Araújo, 2022).

Pleasantness – the aesthetic quality of the urban landscape, including architectural diversity and the presence of green spaces (Alfonzo, 2005).

Utility – which relates to the quality of urban services (Mehta, 2008)

Sensory Pleasure – arising from a harmonious balance between variety and order (Mehta, 2008).

Sense of Belonging – which is nurtured in spaces with historical significance and essential daily functions (Mehta, 2008).

Density – the intensity of urban form, which can enrich or overwhelm social interaction (Sonta and Jiang, 2023).

Several frameworks have been developed to assess multiple dimensions of walkability (Venerandi et al., 2024). Filipe Moura's (Moura et al., 2017) development of the IAAPE (Indicators of Pedestrian Accessibility and Attractiveness in English) represents a significant advancement for evaluating walkability, as it emphasises the importance of accommodating pedestrians alongside other modes of transport (Ferreira, 2017). The Global Walkability Index (Krambeck, 2006) highlights the necessity of robust political support for improving pedestrian infrastructure – a point echoed by Moura through the commitment to the pedestrian mobility concept. Additionally, the discussion of urban legibility stresses the importance of an individual's ability to recognise and structure space for effective navigation (Araújo, 2022). Further complicating the picture, Amor (Amor, 2011) examined the diminishing autonomy in elderly urban mobility, highlighting that a specific urban environment may not offer the same walkability conditions to all citizens. These findings reveal that older people may become isolated, heightening their feelings of vulnerability.

Urban attractiveness, in turn, emerges from the dynamic interplay between environmental factors and individual needs, collectively shaping perceptions of walkability (Buckley et al., 2017). Yet a persistent paradox remains: although active mobility is promoted as a healthy mode of transportation, pedestrians are often the most exposed to air pollution from motorised traffic, especially in densely trafficked areas (Presidência do Conselho de Ministros, 2023). A more detailed comparison of approaches, their advantages and limitations, regarding walkability assessment can be found elsewhere (Li et al., 2023; Fina et al., 2022; Arellana et al., 2020).

Gutiérrez-Alonso (2023) studied the physical environment near a BRT system already operating in Mexico to assess if the stations lack ideal conditions for the population to use the transport system (Gutiérrez-Alonso, 2023). The author used a specific method, based on several quantitative parameters to evaluate walkability, to define five ‘pedestrian quality levels’, assessing discrete features 100 m around each BRT station. Another study, carried out by Munoz-Raskin (2007), evaluated, for urban and suburban areas, the influence of walking distance to BRT stations in Bogota (Colombia) on the urban property values. Vergel-Tovar (2025) explored how built-environment characteristics relate to BRT ridership in the Indian cities of Indore and Ahmedabad. In that study, the data collection protocol was adapted from earlier BRT stop-area assessment approaches developed in Latin American contexts and then applied to the Indian case studies to capture comparable attributes (e.g., pedestrian infrastructure quality, land-use diversity and public facilities). Key factors considered included the quality of pedestrian infrastructure, land-use diversity, public facilities presence and housing provisions to increase residential density (Cervero et al., 2004).

Although the studies above analysed the urban environment around BRT stations, they refer to cities with contexts very different from Coimbra, the city considered in this study. Those cities are located in South America and India, with patterns of planning and physical organisation that contrast sharply with the Portuguese context. Unlike Coimbra, which has a more compact urban form and a distinct historical layout, these cities reflect different cultural, social and infrastructural realities that uniquely shape their development and mobility systems. Moreover, the applied tools by those authors are considerably different from the OS-WALK-EU tool (Fina et al., 2022) used in this paper to assess walkability based on geospatial data, highlighting spatial asymmetries and identifying priority areas for pedestrian-oriented interventions.

Coimbra is a medium-sized city in Portugal that maintained a compact urban form until the twentieth century, after which it underwent significant urban sprawl driven by increased car dependency and the rising value of peripheral land (Coutinho-Rodrigues et al., 2023). Today, the city is preparing to implement the Metrobus system, a bus rapid transit (BRT) network scheduled to start operating in 2025 as the first phase of a full electric mobility system. The two planned lines are expected to reshape mobility patterns by providing high-capacity, sustainable public transport along strategic urban corridors (Metro Mondego, n.d.). In this context, evaluating the walkability of areas surrounding BRT stations is essential to anticipate potential increases in pedestrian trips and to ensure the new infrastructure effectively promotes sustainable urban mobility.

The present study assesses walkability based on geospatial data to address this challenge, highlighting spatial asymmetries and identifying priority areas for pedestrian-oriented interventions. Therefore, the paper contributes to using a previously validated open-source tool developed for European cities (Fina et al., 2022) in a specific context. Although the tool allows users to include local datasets whenever available, which was not the case, the approach used open-source datasets to obtain valuable insights into the interplay between new transit infrastructure and pedestrian accessibility. This approach offers a replicable methodological framework for cities implementing similar mobility systems, even if local datasets are unavailable and gathering them is time-consuming and provides valuable information for the city governance and planning professionals.

2. Methodology

2.1 Overview

Neighbourhood walkability is operationalised with OS-WALK-EU v1.0 (QGIS plugin) on a regular grid, combining four sub-indices: amenity proximity (A), pedestrian-network permeability ( $P_{n}$ ), accessible green/blue ( $G_{i}$ ) and residential density (P_d). The workflow uses open data – OSM for pedestrian networks and POIs (points of interest), Copernicus DEM for topography, and official population statistics – and computes network-based distances via OpenRouteService (ORS) (Statistics Portugal, 2025). Default parameters support cross-city comparability; study-specific values and weights are specified in Sections 2.2 to 2.6. Figure 1 provides a step-by-step overview of the OS-WALK-EU processing chain used in this study, from open-data inputs (OSM POIs and pedestrian network, DEM, land-cover polygons, and census counts) to the computation of the four sub-indices at the grid-cell level. It also shows how the slope-dependent effective walking radius is applied before calculating the shed-based indicators (P_n and G_i), followed by weighted aggregation and normalisation to a 0 to 100 score.

Figure 1.

OS-WALK-EU workflow implemented in this study: data inputs, grid-based computation of A, P_n, G_i, and P_d (with slope-adjusted sheds for P_n and G_i), and final aggregation/normalisation to W_s (0–100).

2.2 Data and preprocessing

OS-WALK-EU evaluates the urban fabric on a regular grid with a default cell size of 500 × 500 m to support cross-city comparability. The grid can be adapted to local conditions; in the case study, a finer 200 × 200 m mesh is used in station corridors to capture local heterogeneity while preserving computational tractability.

OSM POIs are harmonised into OS-WALK-EU v1.0 categories (Retail; Entertainment; Food & Drinks; Civic & Institutional; Office; Sports & Recreation). A routable pedestrian network is prepared from OSM with pedestrian-appropriate attributes and barriers. Green/blue polygons are standardised for overlay and area computations. Population counts are assigned to grid cells for density estimation; sources and spatial allocation procedures are detailed in Section 3. Cell centroids act as origins for ORS routing.

2.3 Parameterisation

OS-WALK-EU provides a compact parameterisation panel in QGIS (Figure 2): input layers are selected on the left, while amenities are weighted on the right. Three groups of settings are available:

Amenity categories (selection, COUNT and relative importance). Up to three amenity categories can be included. For each selected category, the user sets: (i) COUNT $N_{c}$ (number of nearest amenities to consider per category $c$ ), and (ii) a category weight in the range 0.5 to 3. Distances are computed on the ORS network; decay and rank weights (1st/2nd/3rd) are applied to those N nearest amenities.

Search radius ( $R_{0}$ , plugin parameter ‘maxDist’). The initial Euclidean search radius is used to build the network service area and the amenity buffer. This study assumed a default value of 500 m.

Dimension weights (index aggregation). Continuous weights in $[0, 1]$ are assigned to the four sub-indices – A, P_n, G_i, P_d. The chosen weights are used directly in the aggregation; the final score is normalised by the theoretical maximum (Section 2.6). Unless otherwise noted, equal weights are applied to preserve cross-city comparability.

Figure 2.

OS-WALK-EU parameterisation in QGIS used in this study: (a) selection of input layers; (b) configuration of amenity categories (COUNT and category weights), maximum distance (R0), and sub-index weights for aggregation.

The parameter values used in this study are fixed across the analysis area and use the plugin's default settings in the case study (Section 3).

2.4 Slope treatment

Slope is not treated as a standalone indicator in OS-WALK-EU v1.0. Instead, it is modelled as an impedance factor that reduces the effective walking reach used to delineate the pedestrian catchment (‘shed’) before computing shed-based indicators. Let $R_{0}$ be the nominal maximum walking distance (in metres) set in the plugin (in this study, $R_{0} = 500 m$ ). For each grid cell, the mean local slope g (in %) is derived from the DEM and translated into a stepwise reduction factor $k (g)$ .

Table 1 operationalises the slope effect by providing a discrete mapping from the mean cell slope $g$ (%) to a reduction factor $k (g)$ . For each grid cell, g is classified into the corresponding slope interval in Table 1, and the associated k value is retrieved. This discretisation is intended to approximate the reduction in walking speed described by Waldo Tobler's hiking function $v (s) = 6 \exp (| s + 0.05 |)$ (km/h), where s is slope (rise/run), by applying a conservative reduction to the reachable distance rather than explicitly modelling travel time. The slope-adjusted effective walking radius is then defined as:

R_{eff} = k (g) R_{0} .

Table 1.
Slope-dependent reductions to the maximum acceptable walking distance used in OS-WALK-EU v1.0.

g (%) ≤7 >7–15 >15–25 >25–45 >45

k (g) 1.0 0.8 0.6 0.4 0.2

g (%)	≤7	>7–15	>15–25	>25–45	>45
k (g)	1.0	0.8	0.6	0.4	0.2

Using the pedestrian network, a service area is first generated with radius $R_{0}$ and then reduced using $R_{eff}$ , resulting in a slope-adjusted shed $S_{adj}$ . This adjustment affects only the indicators that depend on the shed geometry:

Permeability ( $P_{n}$ ): only the reachable area (numerator) changes because it is computed from $S_{adj}$ , while the reference area remains the full circle with radius $R_{0}$ (denominator $= π R_{0}^{2}$ ).

Accessible green and blue ( $G_{i}$ ): both the accessible green/blue area and the shed area are computed within $S_{adj}$ .

Amenity proximity ( $A$ ) and residential density ( $P_{d}$ ) are unaffected by slope in v1.0: A is derived from network distances to the N nearest amenities (after Euclidean pre-filtering within $R_{0}$ ), and $P_{d}$ depends only on the population allocated to grid cells.

Intuition: Steeper terrain reduces the realistically reachable pedestrian catchment around a cell, penalising network permeability and access to green/blue spaces even when the nominal distance threshold remains $R_{0}$ .

2.5 Indicators to be calculated (per grid cell)

Amenity proximity ( $A$ ; points 1–10). For each cell centroid, candidate amenities are first selected using a Euclidean buffer of radius $R_{0}$ (metres). For each selected amenity category c, OS-WALK-EU computes network distances $d_{c, r}$ (metres) from the centroid to all candidates using ORS routing, retains the $N_{c}$ nearest amenities (where $N_{c}$ is set by the category COUNT parameter), and applies a distance-decay weight:

w (d) = \exp (- \frac{(d - 400)^{2}}{10^{6}}),

where d is in meters (as implemented in OS-WALK-EU v1.0). Within each category, amenities are also rank-weighted (e.g., 1st/2nd/3rd closest), and the category score is combined with a user-defined category weight (range 0.5–3). Category-weighted scores are normalised to a 0 to 100 scale and then discretised to points 1 to 10 using the plugin's fixed thresholds.

Intuition: A increases when multiple relevant destinations are available within short network distances, reflecting how easy it is to satisfy daily needs on foot.

2.
Pedestrian-network permeability ( $P n$ ; points 1–10). A pedestrian network shed is generated and then slope-adjusted (Section 2.4), yielding $S_{adj}$ . Permeability is computed as a PedShed percentage:
$PedShed = \frac{Area (S_{adj})}{π R_{0}^{2}} \times 100 \% .$

This value is mapped to points 1 to 10 using OS-WALK-EU v1.0 ranges.

Intuition: $P_{n}$ is high when the street/path network allows pedestrians to reach most of the theoretical circular catchment (few barriers, good connectivity, many route options). 3.
Accessible green and blue ( $G_{i}$ ; points 1–10). Let G be the total area (m²) of green/blue polygons intersecting the slope-adjusted shed $S_{adj}$ . The green/blue share is:
$G_{i} = \frac{G}{Area (S_{adj})} \times 100 \% .$

The percentage is classified to points 1 to 10 using OS-WALK-EU v1.0 breakpoints (e.g., the highest class corresponds to $G_{i} \geq 22.5 \%$ , and the lowest to $G_{i} < 2.5 \%$ ).

Intuition: $G_{i}$ captures how much green/blue space is realistically accessible within the local pedestrian catchment, which is relevant for comfort, shade, and recreational walking potential. 4.
Residential density ( $P_{d}$ ; points 1–10). Population is assigned to grid cells (Section 3), and density is computed as persons per hectare. Cells are then ranked by deciles across the study area: decile 10 receives 10 points, …, decile 1 receives 1 point.

Intuition: Higher local density tends to support more walkable environments by sustaining services and public transport demand, although it does not guarantee good pedestrian design by itself.
2.6 Aggregation, normalisation and cartographic classes

Let $A, P_{n}, G_{i}, P_{d} \in {1, \dots, 10}$ be the point scores per cell, and let $w_{A}, w_{P_{n}}, w_{G_{i}}, w_{P_{d}}$ be the corresponding non-negative weights set in the plugin (in this study, all were set equal). The composite walkability points are computed as:

W p = A w_{A} + P_{n} w_{P_{n}} + G_{i} w_{G_{i}} + P_{d} w_{P_{d}} .

The theoretical maximum for a given weight set is obtained by assigning 10 points to every sub-index:

W p_{max} = 10 (w_{A} + w_{P n} + w_{G i} + w_{P d}) .

The normalised walkability score is then:

W_{s} = 100 \frac{W p}{W p_{max}} \in [0, 100] .

For cartographic visualisation, $W_{s}$ is displayed using five fixed classes to aid interpretation and cross-area comparison: $\leq 30$ (very low), $> 30 - \leq 45$ (low), $> 45 - \leq 60$ (medium), $> 60 - \leq 75 %$ (high), and $> 75$ (very high).

2.7 Worked example (illustrative)

This example illustrates the calculation steps for one grid cell (values are hypothetical and provided for transparency).

Amenity proximity (A). Assume the plugin configuration includes two categories: supermarkets (COUNT $N = 5$ ) and pharmacies (COUNT $N = 2$ ). If the N nearest network distances are $d = 574$ m for the five supermarkets and $d = 759$ m for the two pharmacies, the decay weights are:

w (574) = \exp (- \frac{(574 - 400)^{2}}{10^{6}}) \approx 0.970, w (759) = \exp (- \frac{(759 - 400)^{2}}{10^{6}}) \approx 0.879.

A simplified (illustrative) sum is $5 \times 0.970 + 2 \times 0.879 = 6.61$ . If the configured theoretical maximum for this setup is 7.0, then $A = (6.61 / 7.0) \times 100 % \approx 94.4 %$ , which maps to 10 points under the v1.0 thresholds.

Slope-adjusted shed. With $R_{0} = 500 m$ and mean slope $g = 30 \%$ , Table 1 selects $k (g) = 0.4$ (class $> 25 - \leq 45 %$ ), so:

R_{eff} = 0.4 \times 500 = 200 m .

Other dimensions (illustrative points). Suppose $P_{n} = 8$ , $G_{i} = 5$ and $P_{d} = 7$ .

Aggregation (equal weights). With $w_{A} = w_{P_{n}} = w_{G_{i}} = w_{P_{d}} = 1$ :

W p = 10 + 8 + 5 + 7 = 30, W p_{max} = 10 (1 + 1 + 1 + 1) = 40, W s = 100 \frac{30}{40} = 75.0.

2.8 Limitations

Validating OS-WALK-EU through correlations with actual pedestrian flows might be unlikely for some urban areas, as walking behaviour also depends on social, lifestyle, and perceptual factors, so the tool should be seen as a preliminary guide for planners rather than a prediction measure of pedestrian flows (Fina et al., 2022).

There might be some limitations of OS-WALK-EU, regarding its reliance on OSM data, which may contain incomplete tagging across different urban areas. Although these data gaps can influence the score of the walkability indicators and, consequently, make it challenging to perform comparisons with great accuracy, the scores are of great utility to urban planning decisions on future changes in the city environment. By combining the systematic methodology of OS-WALK-EU with additional field validation measures, interpreting the quality of walkable spaces around the BRT system may give further insights.

As the BRT system is still under implementation, this study was limited to identifying areas surrounding the corridor that may require attention from city governance and planners. As adopted in previous studies (e.g., (Moura et al., 2017)), assessing users’ perceptions through audit-based walkability surveys will only be possible once the system becomes operational. Such future analyses will also demand more resource-intensive tasks, including pedestrian counts obtained through direct observation or sensor-based monitoring. Moreover, the tool version does not account for different groups of citizens with varying walking abilities, such as older adults, people with disabilities, or children.

3. Case study description and data sources

3.1 General description

Coimbra has around 141,000 inhabitants and public transportation in the urban area has been based on buses. More recently, the bus fleet with fossil fuel engines has been partially replaced by electric buses with batteries. After more than 20 years of advances and setbacks, the public company responsible for developing the Mondego's Mobility System is about to implement the first phase of the system, a BRT with two lines, one that links the city centre to a suburban area, and a second line entirely within the urban area. The system was designed to use a dedicated corridor and will come into operation in 2025, covering the municipalities of Coimbra, Lousã and Miranda do Corvo. The infrastructure is planned to comprise approximately 42 km and 42 stations, and the system is expected to carry around 13 million passengers per year when fully operational (Metro Mondego, n.d.). Figure 3 shows a Geographic Information System (GIS) representation of the Coimbra Metrobus stations in the urban area and the city's location in Portugal. The Hospital Line stations are in blue and the Lousã Line stations in red.

Figure 3.

Coimbra Metrobus stations: hospital line (in blue), urban section of the Lousã line (in red) and city boundary (in yellow).

According to the OS-WALK-EU walkability index, the maximum recommended distance for pedestrian journeys within urban areas is set at 500 metres by default, though it can be adjusted within a range of 250 to 2 500 metres. This flexibility allows for adaptation to different mobility needs, accommodating highly mobile individuals and those with restricted mobility. In this study, a 500-metre radius was considered, representing the average pedestrian. A buffer zone was generated around each station to assess walkability in the vicinity of stations within the Mondego Mobility System, representing the area accessible within a 500-metre walking distance via the existing pedestrian infrastructure. This spatial delineation facilitated the definition of the study area (Figure 4), achieved by overlaying a 200 × 200 m grid. This grid size was chosen as a trade-off between the need for detail and the computational efficiency of the plugin. Some BRT station designations were included to reference the further analysis better.

Figure 4.

Coimbra Metrobus stations (red and blue points) and buffer zone (blue polygon); basemap: OpenStreetMap.

The following section presents the data and parameters (weights) of the OS-WALK-EU plugin (v1.0).

3.2 Data and parameters

3.2.1 Population, pedestrian network and network analyses

The population data is sourced from the 2021 Census dataset provided by Statistics Portugal (Statistics Portugal, 2025), the Portuguese statistical authority. The dataset is based on statistical sections that differ from the utilised grid regarding spatial coverage. Statistical sections are designed for demographic and administrative purposes, whereas the grid used in this study is structured for spatial analysis and walkability assessment. Due to these discrepancies, GIS operations were performed to adapt and allocate population data accurately to the grid cells. Therefore, several GIS operations were conducted to ensure that population data is assigned to grid cells using block-level statistics, which were accessed through a Web Feature Service (WFS) (INE, Inspire, 2025).

ORS was used to compute shortest-path pedestrian network distances between each grid-cell centroid and candidate amenities extracted from OpenStreetMap (OSM). The pedestrian network was derived from OSM walkable links (e.g., footways, pedestrian streets, and streets allowing pedestrian access) while excluding non-walkable segments and applying pedestrian-appropriate constraints (e.g., barriers and access restrictions where tagged). Routing requests used the ORS ‘foot-walking’ profile to ensure that distances reflect realistic pedestrian paths rather than straight-line proximity. As a quality check, a subset of origins was visually inspected to confirm that computed routes follow plausible pedestrian links and do not cross major barriers without mapped crossings; this step reduces the risk of artefacts caused by missing or inconsistent OSM tagging.

3.2.2 Amenities (OS-WALK-EU compliant)

Amenities refer to facilities, attractions, or services that encourage pedestrian movement, including those intended for recreational and leisure walking. OS-WALK-EU provides a predefined classification of amenities available in OSM into six categories, derived from a literature review by the plugin's authors.

The six categories of POIs considered include:

Retail stores – Establishments such as clothing shops, footwear retailers, and opticians, catering to medium- and long-term consumer needs.

Entertainment venues are cultural and recreational facilities, including cinemas, museums, and zoos, that promote leisure and social engagement.

Food-related establishments – Grocery stores, restaurants, and bars, facilitating daily food supply and social interaction.

Civic and institutional facilities: Schools, healthcare services (doctors, pharmacies), and administrative offices that provide essential educational, medical, and civic services.

Office locations – Workspaces such as consultancy firms, coworking spaces, and business offices, represent employment centres near residential areas.

Recreational and sporting facilities – Parks, playgrounds, and sports centres, which contribute to physical well-being and community cohesion.

The data was retrieved from OSM using the QuickOSM plugin in QGIS (Copernicus Land Use Service, 2025). A comprehensive categorisation of OSM tags for these POIs is documented in the ‘wiki’ page (Gitlab, n.d.). The weightings for each category can be adjusted in the graphical user interface (GUI), with non-relevant categories assigned a weight of zero.

3.2.3 Green and blue (water) areas

Green and blue infrastructure data were sourced from the Urban Atlas 2018 dataset, which provides comprehensive land cover information for European cities. The dataset includes major areas such as parks and gardens but may not capture smaller spaces critical for local walkability. Nevertheless, at the scale of analysis, focused on the area surrounding the BRT corridor, the influence of tiny spaces, such as isolated street trees, courtyards, or pocket gardens, on walkability is marginal when compared to larger green areas, whose consideration ensures both methodological consistency and relevance to city planning.

3.2.4 Digital elevation model

As the study area features significant slopes impacting walkability, a digital elevation model (DEM) obtained from the Copernicus Land Use Service (EU-DEM v1.1) (Copernicus Land Use Service, 2025) was incorporated. The impact of the gradient on walkability is determined by calculating its average within each grid cell. Initially set at 500 metres, the maximum acceptable walking distance is gradually reduced as the average slope increases, thereby decreasing the pedestrian catchment area.

3.2.6 Weightings

OS-WALK-EU applies weightings to compute a walkability index on a scale from 0 (not walkable) to 100 (highly walkable). The final score is derived from a weighted sum of individual indicator values. Users can either define custom weightings or adopt the default equal-weight approach.

The presence of local facilities and services significantly shapes pedestrian-friendly urban environments. Many daily walking trips fit within the concept of the ‘visit-live-work triangle’. While individuals choose their residences and workplaces, urban planning strategies and infrastructure development primarily determine other frequented destinations – such as retail hubs, educational institutions, and cultural or leisure sites.

To address this diversity of needs, OS-WALK-EU integrates facilities from six distinct categories into its analysis, as previously described in Section 3.2.3. Each category has a unique significance for residents, and the model allows adjustments to both the weighting and the number of amenities considered. This enables more tailored walkability assessments, reflecting urban populations’ priorities and movement patterns.

Currently, the plugin supports different pedestrian profiles, which are differentiated by age (‘under 17’, ‘between 17 and 30’, ‘between 30 and 70’, ‘over 70’) and socio-economic status (‘low’, ‘medium’, ‘high’). The default settings ‘All’ / ‘All’ compute a general walkability value. These profile options were not available in the version used for this study. At the time of analysis, the weighting approach was based on amenity count and amenity weight, corresponding to the current pedestrian profile ‘P00’ (Gitlab, n.d.).

Indicators within the model can be weighted from 0 to 1. In this study, all indicators were assigned equal weight (1), ensuring uniform importance across the analysis. This decision ensured comparability and methodological neutrality, avoiding introducing subjective judgments about the relative importance of individual dimensions. Walkability is a multidimensional construct in which factors interact in complex ways, and there is currently no widely accepted empirical basis for prioritising one dimension over another. Assigning equal weights, therefore, provides a balanced representation of these dimensions while maintaining simplicity and reproducibility of the index. In future stages of this study, once the BRT system becomes operational, incorporating audit-based walkability surveys and pedestrian counts may provide valuable insights for assigning differentiated weights to the various parameters.

4. Case study results and discussion

4.1 Global results for the area under study

After integrating all relevant data and assigning corresponding weightings, a walkability classification for the study area was generated. This classification comprehensively assesses pedestrian accessibility within the urban environment, particularly near the stations along the Metrobus's lines. The results are visually represented in Figure 5, illustrating the spatial distributions of walkability scores across the surrounding area of the two BRT lines. Because the original formulation does not prescribe a single literature-wide cut-point, this study's classification of the urban areas into high and low walkability uses a threshold of 60 on the OS-WALK-EU 0 to 100 scale, as applied in other studies (Mouratidis, 2021).

Figure 5.

Walkability analysis results.

The assessment reveals that walkability levels along the Mondego Mobility System's two urban lines range from very low (0–30) to very high (75–100), highlighting the spatial variability in pedestrian walkability. A more detailed analysis of each of the lines in the urban area was carried out to assess the walkability of the areas surrounding each of them in greater detail, as presented in the following sections.

As presented in Table 2, the Lousã line (urban section) shows a slightly higher mean score (61.56) than the Hospital line (59.65), while both share the same median (65). Dispersion is greater on the Hospital line (Std. dev. 17.22 vs 15.39), indicating more heterogeneous conditions; the Lousã section reaches a higher maximum (90 vs 85). In both lines, at least half of the cells are ≥ 65, i.e., above the 60-point operational threshold used in the maps below.

Table 2.

Summary statistics of OS-WALK-EU walkability scores (0–100) for 200 × 200 m grid cells within the 500 m network-reach along each Metrobus line.

Metrobus line	Frequency (no. of cells)	Mean	Median	Std. deviation	Min.	Max.	Range
Lousã line (urban section)	952	61.56	65	15.388	12.5	90	77.50
Hospital line	605	59.65	65	17.221	15	85	70.00

4.2 Hospital line

A detailed analysis of the Hospital Line (represented in blue) reveals significant variations in walkability levels across its stations. The assessment indicates that while the areas around some stations are highly walkable, others exhibit poor pedestrian accessibility, fluctuating between very low and very high walkability scores. Overall, the surrounding areas of the Hospital Line demonstrate a generally high level of walkability, as evidenced by the fact that six out of its nine stations score above 60 points on the walkability index. This suggests that a substantial portion of the line benefits from well-integrated pedestrian infrastructure, proximity to key urban amenities, and a favourable built environment. However, the remaining three stations, which score below 60 points, highlight localised pedestrian-use challenges.

One of the primary factors contributing to the high walkability scores along the Hospital Line is the significant presence of urban amenities and green infrastructure. As illustrated in Figure 6, the density of amenities (represented by yellow circles) and the distributions of green spaces (highlighted in pink) have positively influenced the walkability index. Among the best performing stations, Republic Square and Sereia stand out with very high walkability scores, as they are strategically located in areas with a strong presence of green spaces, commercial establishments and public services. Their favourable urban configurations enhance pedestrian movement, making them highly accessible for daily commuting and leisure activities.

Figure 6.

Walkability results for the hospital line (left) and pictures of the surrounding area of Republic Square and hospital stations.

Because the Hospital line will serve the city zone where several public and private hospitals are located, it would be crucial that the walkability score is high in the surrounding area of the BRT line near the hospitals. As Figure 7 shows, the area near the hospital reveals low propensity for walkability, requiring improvement soon. Areas with lower walkability scores tend to lack essential amenities, green spaces and pedestrian-friendly infrastructure, resulting in a less accessible environment for pedestrians. This correlation between walkability scores and the required urban atmosphere underscores the importance of mixed land use and urban design strategies in fostering pedestrian-friendly environments.

Figure 7.

Pictures of the surrounding area of the hospitals with a car-oriented design.

4.3 Lousã line

Figure 8 shows the walkability results for the Lousã line and pictures near the station of Coimbra B (the BRT stop near the rail station). The urban section of the Lousã Line, consisting of 14 stations, exhibits a predominantly high walkability level, with most stations registering scores above 60 points. The centrality of the Lousã Line within Coimbra's urban area significantly contributes to its strong walkability performance. By passing through some of the city's most developed and densely populated areas, this line benefits from its proximity to a wide range of services, educational institutions, and commercial hubs. Despite the high walkability levels, some stations fall below the 60-point threshold, indicating areas where pedestrian infrastructure requires improvement.

Figure 8.

Walkability results for the Lousã line (left) and pictures of the surrounding area of the Coimbra B station (interface with railway).

A notable example is the Coimbra B station, where a fragmented pedestrian network and limited permeability negatively impact the walkability score. As depicted in Figure 8, the lack of continuous and safe pedestrian pathways restricts accessibility, making it difficult for pedestrians to navigate the area efficiently. This reflects broader issues of permeability, defined as the extent to which urban spaces facilitate pedestrian movement and multiple route choices. Areas with low permeability often exhibit discontinuous pavements, physical barriers and poor wayfinding, all of which deter pedestrian mobility and reduce walkability scores (Droin et al., 2024), conditions clearly observed in the surroundings of the Coimbra B station.

4.4 Areas of Coimbra benefiting from the implementation of the BRT

The successful implementation of the Mondego Mobility System, in conjunction with walkability scores above 60, has the potential to significantly improve urban mobility in Coimbra by facilitating pedestrian access to key destinations. This system could enhance public transport integration, encourage sustainable mobility practices and contribute to a more liveable urban environment.

Figure 9 shows the vital areas of Coimbra that will experience positive impacts from implementing the BRT lines. These areas are downtown, the city market, Republic Square, the health infrastructure (hospitals), and the Solum neighbourhood. Nevertheless, this impact might considerably increase if improvement measures for walkability are taken in the surrounding areas of the BRT stations.

Figure 9.

Relevant city areas benefiting from the implementation of the Mondego mobility system.

4.4.1 Hospitals area and the university's campus of health

A particularly strategic area that could benefit from improved walkability and public transport integration is the hospital area and the University of Coimbra health campus. These institutions generate a high volume of daily trips, yet they are currently located in areas with some of the lowest walkability scores within the study area. The dominant presence of private vehicles, limited pedestrian infrastructure and discontinuous pavements severely restricts the pedestrian movement of people using public transport. One of the primary challenges in this area is illegal parking, particularly around the Hospitals, where narrow pavements and obstructed pedestrian paths create significant barriers to movement.

As shown in Figure 9, the existing conditions negatively impact individuals with reduced mobility, further reinforcing the need for comprehensive urban interventions prioritising pedestrian accessibility. Given the hospital users’ profile, many need safe and efficient pedestrian infrastructure. Promoting high-quality walkability solutions is crucial to enhancing the overall user experience and fostering a more inclusive mobility system. Moreover, the surrounding areas significantly lack amenities and support services, combined with a fragmented and low-quality pedestrian network. These factors contribute to this area's lowest walkability level among all those analysed.

This assessment suggests that the area should be prioritised for urban intervention to improve the pedestrian environment, which could encourage greater use of the Mondego Mobility System. Investing in walkability improvements in these critical locations could stimulate broader changes in urban mobility patterns, leading to a more efficient, accessible and sustainable city by integrating pedestrian-friendly solutions with the Metrobus network. Coimbra can enhance public transport adoption, reduce car dependency and create a more dynamic and people-centred urban environment.

4.4.2 Downtown

With walkability levels predominantly above 60, the Downtown area of Coimbra stands to benefit significantly from the introduction of the Metrobus system. The presence of numerous amenities, such as shops, restaurants and various services, can leverage the high walkability of the area and the increased foot traffic generated by public transport users, fostering better-quality urban activity and stimulating the local economy. This area integrates stations from both Metrobus lines, allowing the system to serve diverse population segments with different travel purposes. By enhancing connectivity between key urban spaces, the Metrobus system has the potential to revitalise Coimbra's historic centre and reinforce economic and commercial activities, contributing to a more dynamic and accessible urban environment. Therefore, introducing the BRT line can indirectly enhance walkability by strengthening accessibility and connectivity within and beyond the downtown area. BRT stations will act as anchors of pedestrian movement, attracting activity and encouraging short-distance walking trips to and from stops. By improving access to key destinations and reducing reliance on private vehicles, the BRT corridor fosters a more pedestrian-friendly environment, reinforcing the role of downtown as a high-walkability zone. Moreover, to some extent, the BRT infrastructure will bring urban design improvements, such as better pedestrian areas and safer crossings, indirectly amplifying the walkability conditions.

4.4.3 City market and Republic Square

The city market and the Republic Square areas, served by the hospital line, exhibit high walkability levels. Introducing the Metrobus system allows various businesses and services in the area to benefit economically, as improved public transport accessibility is expected to increase pedestrian activity. The city market is likely to attract a broader customer base, given its strategic location and the enhanced accessibility it will gain with the implementation of the system as the market station will be located just a few meters from the building, which will facilitate convenient access for both residents and visitors, reinforcing the market's role as a commercial hub in the city.

4.4.4 Solum neighbourhood

Solum is a residential neighbourhood with essential sports and recreational facilities, including the city stadium and the municipal swimming pools. With the introduction of the Lousã Line, this area will gain enhanced connectivity with several stations. This expansion will facilitate intra-city travel and regional connections between Coimbra and Lousã, improving accessibility for residents, commuters and visitors.

A key advantage of the Metrobus system in this area is the improved access to the city stadium, which will benefit from a dedicated station. This will allow the Metrobus to serve as a strategic transportation option for large-scale events, such as football matches, concerts and other gatherings, reducing dependence on private vehicles and mitigating traffic congestion during peak events. Furthermore, numerous amenities, primarily concentrated within the existing shopping centre, present another significant benefit.

4.5 Literature-based discussion and urban design implications

Mehta (Mehta, 2008) calculated the walking index by averaging the hourly pedestrian volumes recorded across weekdays and weekends around blocks in the Boston metropolitan area (USA). The study highlights that walking behaviour is shaped by a hierarchy of needs operating at the microscale. Survey and interview results also indicate that even along the same main street, people perceived variations in physical form, land use, and social context, influencing their willingness to walk. The evidence of the study confirms the relevance of safety, environmental comfort, and sensory appeal, while also emphasising the role of social dimensions. These insights expanded the findings by suggesting that factors such as usefulness and a sense of belonging are central to understanding walking needs in addition to Alfonzo's (Alfonzo, 2005) framework.

The present study considered the dimensions of the urban environment defined within OS-Walk-EU, specifically population density in the area surrounding the BRT lines, the street connectivity, the density of amenities, green and blue space availability, and the effect of the slope of walking pathways. Therefore, it gives a different perspective on a walkable environment than the studies above.

Walkability was assessed exclusively through these indicators, using several open-source datasets for evaluation. Through a quantitative approach developed for European cities, the study identified the areas around Coimbra's BRT lines that the municipality should prioritise in future urban environment improvements regarding walkability promotion. Moreover, it established a systematic basis to assess the pedestrian flows from and to the BRT stations after the system starts operating.

As mentioned above, the study did not incorporate microscale attributes, such as pavement conditions, pedestrian crossings, street lighting, or the presence of urban furniture, which, while relevant for pedestrian experience, extend beyond the scope of the present work. Nevertheless, low OS-Walk-EU scores indicate structural weaknesses in pedestrian accessibility, such as connectivity of pedestrian routes, lack of vegetation shade to protect pedestrians against extreme heat events, and measures to reduce reliance on private cars. Urban design responses should prioritise continuous, barrier-free sidewalks, well-marked crosswalks, traffic-calming measures, and adequate street lighting. From a strategic planning point of view, plans to promote mixed-use zoning, implement mobility plans that integrate cycling and walking, and expand access to parks and public spaces are essential to strengthen walkability.

5. Conclusions

Based on the OS-WALK-EU tool, this study examined walkability in the city of Coimbra in the surrounding areas of future Metrobus lines. The analysis revealed that certain areas already exhibit high walkability levels, allowing Metrobus users to comfortably and efficiently access services and amenities on foot, even at some distance from planned routes. However, walkability conditions are notably deficient in other areas, such as the vicinity of the city's public hospitals and the rail station of Coimbra B. In these areas, planners and the city governance should act to capitalise on the benefits of the new mobility paradigm to develop a more sustainable, inclusive, and accessible mobility system that enhances public well-being and urban liveability.

The necessary measures require a multidimensional approach that simultaneously addresses the structural and experiential dimensions of the urban environment. Enhancing permeability and pedestrian infrastructure directly contributes to higher connectivity scores in OS-Walk-EU, reducing travel distances and improving accessibility. Environmental quality is reinforced through noise mitigation measures and shaded green areas, which enhance thermal comfort and urban attractiveness. A diversified land use mix and increased residential density near stations ensure continuous pedestrian activity and support local services. Moreover, traffic calming and limitations on private car access enhance safety and reallocate urban space in favour of pedestrians, reinforcing the surroundings of BRT stations as sustainable and socially vibrant nodes. Urban design efforts should prioritise continuous footways, safe pedestrian crossings, traffic calming measures, shading, and adequate lighting.

The OS-WALK-EU tool, combined with open geospatial datasets such as OSM, Copernicus, and population censuses, supported walkability analysis in Coimbra. As open data is globally accessible, this methodology can be easily applied to other cities. Moreover, it eliminates the need for direct data collection, which is time-consuming and resource intensive.

Footnotes

ORCID iDs

Tomás Pereira

Alexandra Ribeiro

Silvino Capitão

Author contributions

Conceptualisation, A.R, and S.C.; methodology, A.R., and L.P.-S.; investigation, T.P. and A.R.; writing—original draft preparation, T.P.; writing—review and editing, A.R., and S.C., All authors have read and agreed to the published version of the manuscript.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

Abastante

Lami

Riccia

, et al. (2020) Supporting resilient urban planning through walkability assessment. Sustainability 12(19): 8131.

Alfonzo

(2005) To walk or not to walk? The hierarchy of walking needs. Environment and Behavior 37(6): 808–836.

Amaral

(1969) Lisboa - Uma Cidade em Transformação .

Amor

(2011) Percorrendo a (c)idade com idosos: a construção urbana da vulnerabilidade’.

Araújo

(2022) Atributos para Caminhabilidade em Calçadas(ACAC): Ferramenta para Boas Práticas no Desenho e Avaliação do Espaço Público para Pedestres.

Arellana

Saltarín

Larrañaga

, et al. (2020) Urban walkability considering pedestrians’ perceptions of the built environment: A 10-year review and a case study in a medium-sized city in Latin America. Transport Reviews 40(2): 183–203.

Ariffin

RNR

Rahman

NHA

Zahari

(2021) Systematic literature review of walkability and the built environment. Journal of Policy & Governance 01(01): 1–20.

Barros

APBG

(2014) A caminhabilidade sob a ótica das pessoas: o que promove e o que inibe um deslocamento a pé?

Blečić

Congiu

Fancello

, et al. (2020) Planning and design support tools for walkability: A guide for urban analysts. Sustainability 12(11): 4405.

10.

Buckley

Stangl

Guinn

(2017) Why people walk: Modeling foundational and higher order needs based on latent structure. Journal of Urbanism: International Research on Placemaking and Urban Sustainability 10(2): 129–149.

11.

Cervero

Murphy

Ferrel

, et al. (2004) Transit-Oriented Development in the United States: Experiences, Challenges, and Prospects Transit Cooperative Research Program. Washington, D.C: Transportation Research Board TRB. Available: https://nap.nationalacademies.org/catalog/23360/transit-oriented-development-in-the-united-states-experiences-challenges-and-prospects.

12.

Chakrabarti

Shin

(2017) Automobile dependence and physical inactivity: Insights from the California household travel survey. Journal of Transport & Health 6: 262–271.

13.

Copernicus Land Use Service (2025) [Online]. Available: https://plugins.qgis.org/plugins/QuickOSM.

14.

Costa

Neto

GCM

Bertolde

(2017) Urban mobility indexes: A brief review of the literature. Transportation Research Procedia 25: 3645–3655.

15.

Coutinho-Rodrigues

Monteiro

Sousa

, et al. (2023) The impact of city form and active modes of transport on urban mobility energy consumption. In: CEES 2023, 2nd International Conference on Construction, Energy, Environment & Sustainability. Funchal, Portugal. Itecons – Institute for Research and Technological Development in Construction, Energy, Environment and Sustainability, 1–8.

16.

Distefano

Leonardi

(2023) Fostering urban walking: Strategies focused on pedestrian satisfaction. Sustainability 15(24): 16649.

17.

Droin

Wurm

Weigand

, et al. (2024) How does pedestrian permeability vary in and across cities? A fine-grained assessment for all large cities in Germany. Computers, Environment and Urban Systems 110: 102115.

18.

El-Kholei

Yassein

(2022) Professionals’ perceptions for designing vibrant public spaces: Theory and praxis. Ain Shams Engineering Journal 13(5): 101727.

19.

Elsamani

Kajikawa

(2024) Envisioning the future of mobility: A well-being-oriented approach. Sustainability 16(18): 8114.

20.

Ferreira

(2017) Índices de Caminhabilidade Urbana: Aplicação do projeto IAAPE ao planeamento urbano.

21.

Fina

Gerten

Pondi

, et al. (2022) OS-WALK-EU: An open-source tool to assess health-promoting residential walkability of European city structures. Journal of Transport & Health 27: 101486.

22.

Gitlab (n.d.) [Online]. Available: https://gitlab.com/ils-research/os-walk-eu/-/wikis/Predefined-profiles.

23.

Gutiérrez-Alonso

(2023) Evaluación de entornos peatonales en estaciones BRT. Espiral, Revista de Geografías y Ciencias Sociales 5(9): 5–23.

24.

INE, Inspire (2025) Instituto Nacional de Estatística, Lisbon. [Online]. Available: https://inspire.ine.pt/geoserver/BGRI2021/wfs.

25.

Krambeck

(2006) The Global Walability Index.

26.

Yabuki

Fukuda

(2023) Integrating GIS, deep learning, and environmental sensors for multicriteria evaluation of urban street walkability. Landscape and Urban Planning 230: 104603.

27.

AW-T

Houston

(2018) How do compact, accessible, and walkable communities promote gender equality in spatial behavior? Journal of Transport Geography 68: 42–54.

28.

Mehta

(2008) Walkable streets: Pedestrian behavior, perceptions and attitudes. Journal of Urbanism: International Research on Placemaking and Urban Sustainability 1(3): 217–245.

29.

Metro Mondego . (n.d.) Metro Mondego. [Online]. Available: https://www.metromondego.pt/en.

30.

Moura

Cambra

Gonçalves

(2017) Measuring walkability for distinct pedestrian groups with a participatory assessment method: A case study in Lisbon. Landscape and Urban Planning 157: 282–296.

31.

Mouratidis

(2021) Urban planning and quality of life: A review of pathways linking the built environment to subjective well-being. Cities 115: 103229.

32.

Munoz-Raskin

(2007) Walking Accessibility to Bus Rapid Transit in Latin America: Does It Affect Property Values? The Case of Bogota, Colombia.

33.

Ngossaha

Ngouna

Archimède

, et al. (2024) Toward sustainable urban mobility: A multidimensional ontology-based framework for assessment and consensus decision-making using DS-AHP. Sustainability 16(11): 4458.

34.

Presidência do Conselho de Ministros (2023) Resolução do Conselho de Ministros n.^o 67/2023’.

35.

Ros-McDonnell

de-la-Fuente-Aragón

Ros-McDonnell

, et al. (2024) Toward resilient urban design: Pedestrians as an important element of city design. Urban Science 8(2): 65.

36.

Sonta

Jiang

(2023) Rethinking walkability: Exploring the relationship between urban form and neighborhood social cohesion. Sustainable Cities and Society 99: 104903.

37.

Sousa

(2013) A mobilidade automóvel em Portugal. A construção do sistema socio-técnico, 1920–1950.

38.

Statistics Portugal (2025) [Online]. Available: https://tabulador.ine.pt/indicador/?id=0011609&lang=EN.

39.

Szibbo

Church

Miranda

, et al. (2012) Cities and people project: a white paper on human interaction with the built environment. Institute for Environmental Entrepreneurship, 1–67.

40.

Talen

Jeong

(2019) What is the value of ‘main street’? Framing and testing the arguments. Cities 92: 208–218.

41.

Venerandi

Mellen

Romice

, et al. (2024) Walkability indices—the state of the art and future directions: A systematic review. Sustainability 16(16): 6730.

42.

Vergel-Tovar

(2025) The environment around bus rapid transit stops in Indore and Ahmedabad, India. Journal of Urbanism: International Research on Placemaking and Urban Sustainability 18(2): 216–239.