Sage Journals: Discover world-class research

Abstract

Maintaining street vitality, understood as a concentration of human activity, is a key priority for urban agendas, yet predicting and understanding it remains challenging. This paper examines street vitality in 10 European cities using functional density as a proxy, measured as the number of unique points of interest (POI) classes per street segment length. Leveraging OpenStreetMap (OSM) data and a spatial lag regression model with a log-transformed dependent variable, the analysis accounts for spatial dependencies while examining the determinants of functional density across cities and within street segments characterised by high functional density. Findings show that morphological factors influence functional density more consistently than transport-related ones. Higher clustering of functional density in city-wide models indicates the need for strategic planning around vitality, whereas the weaker clustering observed in models focused on streets with high functional density indicates greater consideration needed for local street-level attributes. Long street segments appear to enhance functional density at the city level, while compact urban forms support vitality in already vibrant areas. Commercial density enhances functional density city-wide, while residential presence is more significant in streets with high functional density. Transport infrastructure explains better street vitality clustering at the city level, but its role weakens among other variables in already dense areas.

Keywords

street vitality functional density POI spatial modelling

Introduction

Street vitality reflects the presence of people and the diversity of urban functions; therefore, it refers to a high concentration of human activity (Eraydin and Özatağan, 2021; Hao et al., 2015; Xia et al., 2022). Street vitality serves as an indicator of human presence, community life, and multifunctional use (Gehl, 2010; Jacobs, 1961; Mehta, 2007; Montgomery, 1998), and under the broader concept of urban vitality, it has been considered a marker of economic activity and a key driver of sustainable urban development (Ratcliffe and Flanagan, 2004; Ravenscroft, 2000; Tomalin, 1997). In recent years, the rise of data-driven urban research has further expanded the understanding of street vitality by incorporating real-time digital footprints, such as smartphone activity, mobile network data, user-generated content, pedestrian mobility patterns, taxi flows, and night-time lighting (Han et al., 2023; Wang et al., 2024; Xia et al., 2022).

Street vitality is closely related to concepts such as liveable and walkable streets, although each has its own sphere of understanding. Liveability reflects a city’s overall quality of life and it can be experienced at the street level wherever people live, move, and interact daily (Appleyard and Appleyard, 2020; Istrate and Chen, 2022; Istrate et al., 2021; Sanders et al., 2015), while street vitality concentrates in city centres or along the main streets of residential neighbourhoods (Ravenscroft, 2000; Xia et al., 2022). While street vitality contributes to the broader concept of liveability, it does not fully encompass it. Similarly, walkability supports vitality by addressing how easy and safe is to walk in an area, being often assessed through convenient pedestrian infrastructure like sidewalks and crossings (Balsas, 2019; Frank et al., 2010). However, high pedestrian counts alone do not necessarily indicate street vitality – for example, pedestrians might be present due to necessity, proximity to a transit hub, or simply passing through, without meaningful social or commercial activity (Appleyard and Appleyard, 2020; Harvey and Aultman-Hall, 2016). While convenient walking conditions may facilitate lively streetscapes, street vitality encompasses primarily the diversity of human activities in public spaces (Gehl, 2010; Jacobs, 1961). Another more recent concept of social sustainability focuses on long-term social interaction, a sense of place, social participation, safety, social equity, and neighbourhood satisfaction (Kyttä et al., 2016; Larimian and Sadeghi, 2021). However, street vitality highlights the dynamic, everyday activity on streets, driven by the presence of people, social interactions, economic activity, and the quality of the physical environment altogether (Park et al., 2013; Xia et al., 2022).

Studying street vitality is challenging due to the concept’s ambiguity, making case studies the common approach. While this allows obtaining rich local data, the context-specific insights are hard to generalise, and some scholars consider them insufficient for city-wide analysis or replicable urban planning interventions (Xia et al., 2022). The central research question driving this investigation is: What factors consistently influence functional density – and, by extrapolation, street vitality – and how does area sampling affect their interpretation?

This study focuses on functional or POI (points of interest) density as a commonly used proxy for street vitality (Tu et al., 2020; Yue et al., 2017). A high functional density indicates a well-developed area with a diverse array of services and amenities, showcasing the vibrancy and appeal of the space (Barreca et al., 2020; Tu et al., 2020). In contrast to many studies that use land-use mix indices, which focus on broad zoning categories like residential, commercial, and industrial land uses, POI density captures the detailed variety of services and facilities within neighbourhoods, and at the street level (Long and Huang, 2019; Wang et al., 2024).

The factors most frequently examined in recent studies of street vitality include morphological and transport-related aspects. Morphological factors relate to the physical structure and land-use patterns in urban settings (Lu et al., 2019a; Wangbao, 2022). For example, building density plays a key role, as higher-density areas support more intense human activity, fostering diverse uses and interactions (Han et al., 2023; Niu et al., 2022). Street vitality is influenced by street blocks or segment lengths, as several authors, starting from Jane Jacobs (1961), showed that shorter blocks tend to promote higher pedestrian activity, whereas larger blocks may impede movement and diminish vitality (Fang et al., 2021; Long and Huang, 2019; Wang et al., 2021). Streets lined with shops, cafes, and restaurants attract a steady flow of people, fostering social interaction and economic activity (Jiang et al., 2022; Sung and Lee, 2015). Office spaces contribute to vitality by generating activity during working hours, with workers moving between their offices and nearby amenities (Huang et al., 2023; Yue and Zhu, 2019). Residential buildings add another layer to street vitality by providing a base population that sustains activity during non-working hours, particularly in the evenings and on weekends (Park et al., 2013; Zhang et al., 2021). The presence of residents ensures that streets are used throughout the day, not just during business hours (Lu et al., 2019b; Yang et al., 2021). Other building functions may influence vitality differently. Sports buildings generate periodic bursts of activity as they attract crowds for events or regular visits (Austrian and Rosentraub, 2002; Klinmalai and Kaewlai, 2023). Industrial buildings, in contrast, tend to have a more complex relationship with street vitality. They can either contribute to economic activity and employment or create low-traffic zones, depending on their location and integration into the urban fabric (Tu et al., 2020; Vukmirović and Nikolić, 2023).

Transport-related variables reflect the accessibility provided by various modes of transportation. Public transport access, represented by bus stop density, tram stop density, and railway station density, plays a central role in shaping street vitality (Yu et al., 2022; Zheng et al., 2023). Streets with higher densities of public transport stops typically experience more foot traffic, as they are more accessible to a larger number of people (Jiang et al., 2022; Zhang et al., 2021). Fixed-rail systems like trams and railways are especially important in dense urban environments, as they attract higher volumes of passengers, supporting transit-oriented development and enhancing street vibrancy (Han et al., 2023; Zacharias, 2020). The availability of taxis responds to street-level activity by ensuring that people have multiple options to access urban areas (Park et al., 2013; Yuan and Chen, 2021). While parking facilities are often associated with increased car use, they still play an important role in making streets accessible to visitors who rely on private vehicles (Shiftan and Burd-Eden, 2001). However, excessive reliance on parking spaces can affect walkability, which is important for supporting vibrant urban areas (Still and Simmonds, 2000). Traffic signals and crossings are essential elements of urban infrastructure that facilitate both vehicular and pedestrian movement (Baptista Neto and Barbosa, 2016; Gómez-Varo et al., 2022).

In summary, the morphological factors focus on the physical structure and land-use patterns that encourage diverse activities, while transport-related factors emphasise the importance of accessibility and movement in shaping the vitality of streets. It is assumed that these two categories of factors, when working in tandem, create the conditions for street vitality, supporting social interaction, movement, and triggering economic activity. Both categories are considered in the conceptual framework of this study (Figure S1 in Supplemental Material).¹

The paper will detail the methodologies employed, describe the data and study areas used, present the results of the analysis, and conclude with a discussion on the implications of these findings for urban planning and development. By addressing the factors that enhance or detract from street vitality, this research will help inform strategies that foster vibrant urban environments.

Materials and methods

Sample

The study area comprises 10 European cities participating in the REALLOCATE project, funded by the European Union, with different population sizes, but all with the goal of making their streets more pedestrian-friendly and attractive (Table S1 in the Supplemental Material). This selection of cities offers diverse urban forms and functions, allowing for comparison across contexts. It helps identify consistent factors shaping street vitality and informs planning strategies to support vibrant streets.

Data analysis and data sources

The data analysis relies on a quantitative approach, meaning it focuses on numerical data and statistics to study the relationships between variables like functional density, building density, and other morphological and transport infrastructure elements. The key process involved creating a street network dataset by merging streets based on their names and then segmenting it at intersections, resulting in individual street segments with unique ID. The street segments were analysed using spatial and statistical models to identify patterns and factors influencing street vitality across different cities. A spatial regression model was applied to examine the relationship between the spatial distribution of vitality and the built environment characteristics of the streets.

Area, street density, number of buildings, median building density, street length, and number of POI for every city were calculated in QGIS 3.28.12-Firenze from open-source data. The primary data for this research comes from OpenStreetMap (OSM), a free, open-source map platform. Specifically, it includes information on street networks, buildings, POI, and transport infrastructure. In August 2024, the data was downloaded via Geofabric.de, a service that packages OSM data for specific regions. The city boundaries, which help define the area of study, were also gathered using the tool ‘Open OSM’ in QGIS, a geographic information system software. In recent years, OSM data have become the most widely used form of open geospatial data across various interdisciplinary fields, especially in urban studies (Wang et al., 2022; Zhang et al., 2022). The value of using OSM in this study lies in its ease of data collection across the 10 cities, making it cost-effective and less time-consuming to replicate, because OSM data are freely accessible to all, offer global coverage, include a wide range of mapping features such as land cover/use, roads, and buildings, and are updated in real-time (Grinberger et al., 2022; Zacharopoulou et al., 2021). Additionally, there is a relatively high completeness of OSM estimated for Europe (Herfort et al., 2023).

Vitality proxy

A total of 134 POI classes were identified in the sample (Figure S2 in the Supplemental Material). The distribution of these classes across cities is relatively uniform, averaging approximately 100 ± 20 POI classes per city. However, the uneven distribution of POI data for certain cities, like the high concentration of recycling points in Tampere and Utrecht, and tourist information services in Gothenburg and Heidelberg, can distort vitality measures. To address this issue, functional density was calculated based on the number of unique classes per street segment. Additionally, certain POI classes, such as ‘bench’, ‘camera_surveillance’, ‘waste_basket’, and ‘tower’ (public lighting in industrial zones) were excluded, because they represent microscale infrastructure (urban furniture, which could have a temporary character) rather than functional destinations that contribute consistently to street vitality. Unlike amenities such as shops, restaurants, or recycling services, these microscale elements do not directly indicate economic activity, nor street function that attracts consistent human activity in urban settings, which are key components when considering functional density as a proxy for street vitality. Additionally, their presence is often regulated by urban design standards rather than emerging as a reflection of street vibrancy.

Factors

This study employed 16 indicators derived from morphological and transport-related factors as independent variables (see conceptual framework, Figure S1 in Supplemental Material). Morphological indicators encompass length, building density, and the density of various building categories per street segment length. For every street segment, the length in metres was calculated in QGIS. Building density was determined by dividing the ground floor area by the area of a 50-m street buffer. The densities for commercial, office, residential, individual housing, sports centres, and industrial buildings were calculated by dividing the number of buildings within 50-metre street buffers by the length of the street segments (Figure S3 in the Supplemental Material). A 50-metre buffer was applied because previous research has demonstrated this is the immediate area of influence on each side from the street axis (Oliver et al., 2007), hence capturing the direct influence of a POI on a specific street segment. For the commercial density, data from commercial, mall, leisure, retail, and supermarket building types were combined. The residential density included data from both residential buildings and apartments. For individual housing, data from allotment houses, semi-detached houses, and detached houses were merged. The sports density was calculated using data from sports buildings, sports halls, and sports centres. Finally, the industrial density was derived from industrial and manufacturing building types.

The transport-related indicators comprise the densities of bus stops, tram stops, taxi stands, railway stations, traffic infrastructure, and parking. These densities were calculated by dividing the number of objects by the length of the street segments. The traffic infrastructure density combines data on crossings and traffic signals.

Spatial regression analysis

To examine the factors of street vitality, a spatial lag logarithmic regression model was developed using Python libraries ‘libpysal’ and ‘spreg’ in Google.Colab (Rey and Anselin, 2007). The spatial lag model explicitly accounts for the influence that nearby locations exert on one another (i.e., including neighbouring values of the dependent variable as an explanatory variable in the regression), capturing the inherent spatial relationships that traditional regression models may overlook. The logarithm of the dependent variable (functional density) was taken to handle skewed data distribution. The spatial weights matrix was constructed using a k-nearest neighbours (KNN) approach with k = 4 and row-standardised to ensure comparability across units.

The model is represented by the following equation:

Log (FD) = B 0 + B 1 \cdot BD + B 2 \cdot Length + B 3 \cdot BsD + B 4 \cdot TsD + B 5 \cdot TD + B 6 \cdot RailD + B 7 \cdot PD + B 8 \cdot TiD + B 9 \cdot CD + B 10 \cdot OD + B 11 \cdot InD + B 12 \cdot RD + B 13 \cdot SD + B 14 \cdot ID + B 15 \cdot W_V + E

where:

FD	Functional density (log-transformed)
B0	Intercept
B1,B2,...B15	Coefficients for independent variables
BD	Building density
Length	Length of street segment in metres
BsD	Density of bus stops
TsD	Density of tram stops
TD	Density of taxi stands
RailD	Density of railway stops
PD	Density of parking
TiD	Density of transport infrastructure
CD	Density of commercial buildings
OD	Density of office buildings
IhD	Density of individual housing
RD	Density of residential houses
SD	Density of sports buildings
ID	Density of industrial buildings
W_V	Spatial lag
E	Error term, representing the variation not explained by the model

Results

When examining the spatial patterns of functional density – the street vitality proxy in this study – the areas of concentration are clearly visible in the central parts of the cities (Figure S4 in the Supplemental Material). The city model’s performance ranges from 0.32 in Heidelberg to 0.41 in Budapest, and the high-functional-density model’s performance ranges from 0.2 in Warsaw to 0.35 in Utrecht (Tables S2 and S3 in the Supplemental Material). Model diagnostics using Moran’s I on residuals demonstrate that the spatial regression framework captures a substantial share of spatial structure in both city-wide and high-functional-density street models. In city-wide models, significant negative spatial autocorrelation (e.g. Budapest: −0.23, Barcelona: −0.22; all p < 0.001) suggests that the models effectively reduce spatial clustering and instead reveal more dispersed patterns, likely reflecting nuanced urban heterogeneity. In models focused on the top 25% of streets by functional density, results are more differentiated: in some cities, such as Utrecht (−0.01, p = 0.2) and Heidelberg (0.03, p = 0.1), spatial dependence is no longer significant, indicating strong model performance. In others, such as Barcelona (0.23), Warsaw (0.1), and Bologna (0.1), moderate positive spatial autocorrelation persists (p < 0.001), suggesting locally clustered factors that may merit further exploration. These findings affirm the robustness of the modelling approach across diverse urban contexts, while also highlighting the value of future refinements.

City-wide models

Among the morphological factors, the street segment length showed consistency in all models. As such, street segment length is positively associated with functional density in all cities, with the smallest association in Zagreb and the biggest association in Lyon. A 1-metre increase in street segment length leads to an increase in functional density ranging from 0.001 to 0.007 log units, corresponding to a 0.1% to 0.7% rise in POI density per metre (Table 1).

Table 1.

Regression coefficients^a,b for city-wide models.

Variable	Barcelona	Budapest	Bologna	Gothenburg	Heidelberg	Lyon	Tampere	Utrecht	Warsaw	Zagreb
Constant	−0.62	−0.59	0.00	−1.47	−0.47	−2.47	−4.79	−1.15	−4.01	−2.82
Building density	−0.14	−0.07	−0.16	2.40	−0.48	−0.06	0.21	−0.03	0.07	0.09
Length, m	0.005	0.004	0.002	0.001	0.002	0.007	0.003	0.004	0.004	0.001
Density of individual housing	−20.28	−2.34	32.71	−3.97	−4.98	−13.52	−4.63	4.56	n/a	−1.47
Density of residential houses	−0.17	0.61	0.28	2.59	0.42	3.44	19.00	−0.10	−1.06	3.04
Density of commercial buildings	0.11	14.33	28.70	9.39	15.93	11.72	31.94	3.10	−9.54	12.01
Density of industrial buildings	−5.12	−4.42	1.47	−1.48	−19.41	1.87	−1.63	0.57	n/a	−1.31
Density of sports buildings	−5.64	25.86	8.32	48.14	2.52	23.79	5.26	5.17	n/a	−22.59
Density of office buildings	1.70	6.80	7.14	−0.28	25.57	−5.16	19.95	−1.14	n/a	54.60
Density of taxi stands	5.30	2.75	8.67	86.15	85.68	23.32	37.16	258.68	n/a	48.93
Density of railway stops	17.22	12.08	40.55	40.76	120.85	1.10	−33.10	283.92	n/a	5.43
Density of bus stops	3.56	17.32	23.95	4.33	2.89	4.46	22.77	2.16	n/a	20.71
Density of tram stops	41.05	−13.44	n/a	1.05	20.08	14.66	24.80	−19.70	n/a	4.09
Density of parking	0.72	32.19	3.28	13.19	−0.17	62.60	75.77	0.39	n/a	104.06
Density of transport infrastructure	−2.44	−1.31	−4.28	0.40	−2.00	−2.63	−2.04	−1.13	84.64	8.07
Spatial lag^c	0.91	0.98	1.00	0.88	0.94	0.70	0.65	0.91	0.64	0.76

^aEmboldened coefficients are statistically significant, with p < 0.05.

^bn/a denotes factors not considered in a specific city model for stability.

^cSpatial lag refers to the influence of neighbouring street segments in the statistical model.

Building density shows mixed effects across city models. It has a strong positive association with functional density in Gothenburg, Tampere, Zagreb, and Warsaw. For example, in Gothenburg, a 10% increase in building density is associated with a 27% increase in POI density per metre, based on the coefficient of 2.4 (calculated as 100 × (e^{^}(0.10 × 2.4) − 1)). Conversely, a significant negative effect appears in Barcelona, Bologna, Budapest, Lyon, Utrecht, and Heidelberg, where a 10% increase in building density explains a 0.3% to 3% decrease in functional density. Individual housing density is negatively associated with POI density in five cities, with each additional house per 100 m reducing functional density by 2% to 18%. Industrial density explains a 5% decrease in POI density per additional building per 100 m in Barcelona. Residential building density is positively associated with POI density in Gothenburg, Lyon, Tampere, and Zagreb, with 1 building per 100 m linked to a 3% to 21% increase in POI density. Commercial building density is positively associated with POI density in eight cities, with a 3% to 38% increase for each commercial building per 100 m. Office building density is associated with a 7% to 73% increase in POI density for each additional building per 100 m in Budapest, Heidelberg, Tampere, and Zagreb. Sports building density shows a strong positive association in Gothenburg, with each additional building per 100 m linked to a 62% increase in POI density.

Key transport-related factors associated with functional density across the 10 case-study cities are summarised in Table 1. Bus stop density is positively associated with functional density in five cities, with 1 stop per 100 m linked to a 4% to 27% increase in POI density. Tram stop density is negatively associated with functional density in Budapest and Utrecht, with a 13% to 18% decrease for every 1 stop per 100 m. Railway stop density explains a 33% increase in functional density in Utrecht for every 1 stop per 1000 m. Taxi stand density explains a 5% to 30% increase in functional density in four cities for 1 stand per 1000 m. Parking density explains a 3% to 11% increase in functional density for each additional parking per 100 m in Budapest and Zagreb. Traffic infrastructure density is negatively associated with functional density in Barcelona, Bologna, and Lyon (2% to 4% decrease), but positively associated with functional density in Zagreb (8% increase) and Warsaw (up to 133% increase, meaning that any additional crossing or traffic light explains an increase from average 2 to 3 POI per 100 m). This suggests clustering near POI hotspots.

Spatial clustering (spatial lag) of functional density is significant in all cities, with the strongest association in Bologna.

High-functional-density models

High-functional-density models focused on streets in the top 25% (top-quartile) of functional density (Table 2). Street segment length is the only consistently significant indicator across all cities, with each additional metre linked to a 0.1% to 0.4% decrease in functional density per metre – smallest in Tampere, largest in Bologna (Table S3 in Supplemental Material). Among morphological factors, overall building density and the density of buildings by function showed significant associations with functional (POI) density. Building density had a positive association in several cities – including Barcelona, Heidelberg, Budapest, and Lyon – with the strongest association in Bologna, where a 10% increase explains a 0.007% to 1.4% rise in functional density. Individual housing density is negatively associated with functional density in Heidelberg, Lyon, and Tampere, with each additional building per 100 m linked to a 3% to 5% decrease in POI density. Industrial density is associated with a 3% decrease in POI density per additional building per 100 m in Lyon, but a 2% increase in Utrecht. Residential building density explains a 0.3% to 6% increase in functional density for each additional building per 100 m across eight cities, including Barcelona, Budapest, and Lyon. Commercial building density explains a 0.8% to 3% increase in POI density for each additional building per 100 m in five cities, while office building density explains a 1% to 5% increase in POI density for each additional building per 100 m in four cities. Sports building density in Budapest shows the strongest effect, explaining an increase in functional density by 10% for each additional building per 100 m.

Table 2.

Regression coefficients^a,b for high-functional-density models.

Variable	Barcelona	Budapest	Bologna	Gothenburg	Heidelberg	Lyon	Tampere	Utrecht	Warsaw	Zagreb
Constant	−1.82	−1.83	−1.61	−1.58	−1.58	−1.74	−1.00	−1.29	−2.09	−2.13
Building density	0.01	0.04	0.09	0.00	0.14	0.03	0.02	0.02	0.03	0.06
Length, m	−0.004	−0.002	−0.004	−0.002	−0.003	−0.003	−0.001	−0.003	−0.002	−0.002
Density of individual housing	−13.15	−0.06	−19.85	−0.44	−2.58	−4.65	−4.93	7.59	n/a	−0.72
Density of residential houses	0.30	1.51	0.74	1.23	0.39	0.56	0.85	−0.06	5.66	0.68
Density of commercial buildings	0.84	3.24	−1.20	2.44	1.51	0.78	1.02	0.81	10.98	3.43
Density of industrial buildings	−1.69	−4.36	−0.70	−3.07	1.66	−3.27	2.02	1.49	n/a	0.56
Density of sports buildings	−18.72	9.85	0.10	−1.87	n/a	−16.92	20.03	−34.74	n/a	14.31
Density of office buildings	1.03	1.26	2.16	1.05	−1.72	−1.26	4.35	0.39	17.31	4.95
Density of taxi stands	1.03	6.19	11.54	−2.45	−0.75	2.63	4.09	18.58	n/a	−0.39
Density of railway stops	1.29	5.74	n/a	−4.99	5.97	4.81	31.23	5.00	n/a	n/a
Density of bus stops	0.21	1.97	−0.67	−1.69	0.39	1.96	−1.05	−1.52	−29.10	−0.68
Density of tram stops	−3.70	−0.05	n/a	1.66	−0.12	0.67	1.19	−3.46	n/a	1.80
Density of parking	3.35	−0.75	3.58	−26.06	0.76	5.36	−0.83	n/a	n/a	8.63
Density of transport infrastructure	0.28	0.73	−0.63	−0.52	−1.05	0.25	−0.12	0.02	1.54	0.14
Spatial lag^c	0.07	0.31	0.29	0.35	0.31	0.07	0.66	0.45	0.09	0.19

^aEmboldened coefficients are statistically significant, with p < 0.05.

^bn/a denotes factors not considered in a specific city model for stability.

^cSpatial lag refers to the influence of neighbouring street segments in the statistical model.

Among transport indicators, bus stop density is associated with a 2% increase in POI density for each additional stop per 100 m in Budapest, Gothenburg, and Lyon. Railway stop density explains a 0.5% to 3% increase in POI density for each stop per 1000 m in Budapest, Lyon, and Tampere. Taxi stand density explains increased functional density by 0.6% to 1% for each stand per 1000 m in Budapest and Bologna. Each additional parking per 100 m is associated with a 3% increase in POI density in both Barcelona and Bologna. Traffic infrastructure density also explains a functional density increase of 0.3% to 0.7% for each stand per 100 m in Barcelona and Budapest. Spatial clustering of functional density (spatial lag) appears in six cities, with the strongest in Tampere (Figure S5 in the Supplemental Material) (Galaktionova 2024).

Discussion

Key highlights from the analysis

This research aimed to identify the factors that consistently show influence on functional density, which was used as a street vitality proxy, and how area sampling might affect the interpretations of the results for street vitality.

Analysis of city models and high-functional-density models indicates that functional density is more consistently linked to morphological factors than transport-related ones. Additionally, area sampling significantly influences model outcomes, as results differ between models using city-wide data and those focusing solely on high-functional-density streets. City-wide models show strong spatial dependence, meaning that functional density clusters, as lively streets tend to be near other lively streets. In high-functional-density streets, spatial dependence weakens, suggesting that within already vibrant areas, spatial clustering is not as strong as other factors. The higher performance of city models suggests that data-driven analysis captures better the main influencing factors at the city level, while at the high-functional-density street level, it becomes less straightforward, with more complex interactions.

Some contrasting results also emerged in the two models. In top-quartile areas, shorter street segments support more POI, reinforcing the idea that short blocks contribute to vitality (Gómez-Varo et al., 2022; Jacobs, 1961; Sung and Lee, 2015). However, when considering city-wide data, the opposite trend emerges, as longer street segments (i.e., major arteries) accumulate more POI, although the quality in liveliness might not equate the former. Densely built areas generally positively correlate with high functional density, aligning with established urban theories (Delclòs-Alió et al., 2019; Huang et al., 2023). However, in city-wide models, a negative correlation emerges in six cities, suggesting that higher building density does not always translate to greater functional density, which other authors previously explained based on zoning regulations and monofunctional land uses lacking functional diversity (Li and Pan, 2023; Lu et al., 2019a). The residential density as well as the commercial density are mostly consistent indicators of high functional density, supporting other findings about residential significance (Yang et al., 2022; Yu et al., 2022) and commercial importance (Tang et al., 2018; Zhu et al., 2021). Individual housing density (usually low-density areas) is consistently negatively associated with high functional density. The relationship between industrial buildings’ density and functional density varies between cities and models. In the city model of Barcelona and the top-quartile model for Lyon, industrial density shows a significant negative association with functional density, suggesting that active industrial zones may detract from other functional uses or inhibit vibrancy in these cities. However, in the top-quartile model for Utrecht, industrial density positively correlates with functional density, possibly reflecting the successful adaptation of former industrial sites into lively mixed-use spaces (Radziszewska-Zielina et al., 2022; Vukmirović and Nikolić, 2023). Previous assumptions about the positive effects of sports amenities (Austrian and Rosentraub, 2002; Graham et al., 2023; Klinmalai and Kaewlai, 2023) are reinforced by the significance of sports facilities in explaining high functional density in the top-quartile model of Budapest and the city model of Gothenburg.

Transport-related variables have contrasting, or sometimes non-significant associations with functional density across cities in both types of models. If the data for the whole city is taken into consideration, transport infrastructure plays a significant role in clustering functional density, whereas in areas that already have high functional density, transport infrastructure’s impact becomes weaker. As such, bus stops, railway stations, and taxi stands have strong positive effects on functional density at the city level, corroborating previous findings (Priemus and Konings, 2000, 2001; Yu et al., 2022), but their influence diminishes in already functionally-dense areas due to other contributing factors to vitality.

Limitations

One of the main challenges with OSM data is its inconsistent quality and completeness, which can vary significantly across different regions. OSM relies on crowdsourced data contributions, meaning that the coverage and details of POI are often subject to the involvement of local mappers. In cities with a high number of contributors, OSM data tends to be more complete and accurate, whereas, in less developed areas or regions with lower participation, like Warsaw, data can be sparse or outdated. This variation can lead to unequal representation of urban areas in the analysis, with some streets appearing less vital simply because POI data is incomplete. The OSM platform exhibits biases toward certain types of POI that are more likely to be mapped by contributors. This bias can skew analyses, potentially overlooking facets of street vitality that stem from informal, temporary, or less visible uses of urban space which are not mapped.

Nonetheless, the POI distribution (i.e., functional density), although a sole proxy in this study, provides an understanding of activity and diversity at the street level, which in correlation to morphological and transport-related factors, provides valuable insights into street vitality. While these variables alone cannot capture the full complexity of vitality, an important contribution we make consists in highlighting an area sampling issue in vitality studies: commonly used city-wide models may overestimate vitality, especially in low-vitality areas, regardless of the proxy used (functional density in our study, or potentially smartphone data, social media, etc., in other studies). Future research should probe this further, by using advanced statistical or machine learning models, along with incorporating multi-city or longitudinal comparisons with expanded variable sets.

Other limitations consist of how functional density may not always align with perceived vibrancy (Yue et al., 2017), besides overlooking factors like pedestrian flow, accessibility, or microscale urban design. Street-level data, such as pedestrian counts and business activity, could offer more nuanced and contextual insights. Similarly, including socio-economic variables (e.g., income levels and demographic composition) alongside stakeholder input could offer additional actionable insights into contextual drivers of vitality in future research. The use of these variables could improve model accuracy and reduce potential omitted variable bias. Their absence may leave unexplained variance in the models, contribute to residual spatial autocorrelation, and limit the interpretability of street-level differences across diverse urban contexts. Nevertheless, while recognising the value of city-specific analysis, this study has focused on identifying consistent patterns between street vitality and physical characteristics to reveal generalisable trends (presented above). Future research can explore context-dependent conditions to explain deviations from generalised patterns.

Key policy implications

Policies aimed at enhancing street vitality should be tailored to two distinct contexts: high-functional-density areas versus the city as a whole. At the city level, the strong spatial clustering of functional density suggests that neighbouring areas significantly influence street vitality, emphasising the need for strategic, rather than isolated, planning strategies. However, in streets with high functional density, this spatial dependence weakens, indicating that street-level attributes play a more decisive role than broader neighbourhood effects, requiring targeted, street-specific interventions instead of a one-size-fits-all regulatory approach.

The relationship between street length and functional density further supports this dual strategy. At the city scale, the longer, well-connected corridors enhance functional density overall (i.e., especially outside city centres), whereas, in high-functional-density areas, compact urban forms are more effective in sustaining street vitality. Similarly, commercial density dominates as a driver of functional density at the city level, while in high-functional-density areas, a stronger residential presence is more consistently linked to urban vibrancy.

Conclusion

This study focuses on 10 European cities and investigates functional density calculated as the number of unique POI classes per street length, representing the vitality proxy in this study, using data collected from OpenStreetMap. Street networks were segmented based on intersections, and 134 classes of POI data were used. A spatial lag regression model with a log-transformed dependent variable was employed to account for the spatial dependence between areas, helping to explore the correlation between these built environment factors and the understanding of street vitality. Starting from the hypothesis that functional density can offer relevant insight into street vitality when used as its proxy, this research has created two models with different area samples to compare changes in functional density.

The spatial dynamics of vitality, modelled through the functional density proxy, align with expectations of street vitality clustering across the city. The results highlight sampling-dependent relationships between urban morphological factors and functional density. City-wide models emphasise the role of commercial presence, long street segments that are likely to acquire more POI, along with neighbouring influence in shaping functional density. By contrast, in models on high-functional-density streets (top-quartile), shorter blocks with residential presence explain increases in street vitality and therefore require contextual design and planning. Future research should additionally explore microscale factors alongside morphological and transport-related elements, particularly considering their varying effects in high-vitality areas. Urban planners should account for these nuances when designing interventions to enhance vitality, as outlined in more detail in the key policy implications in the Discussion section.

Supplemental Material

Supplemental Material - Assessing street vitality using functional density as a proxy

Supplemental Material for Assessing street vitality using functional density as a proxy by Anastasiia Galaktionova and Aura-Luciana Istrate in Environment and Planning B: Urban Analytics and City Science.

Footnotes

Acknowledgments

We would like to express thanks to Dr Tiago Tamagusko for his insightful feedback in the initial or ongoing stages of this research.

ORCID iDs

Anastasiia Galaktionova

Aura-Luciana Istrate

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research is developed as part of the REALLOCATE project, funded by the European Commission, grant number 101103924.

Declaration of conflicting interests

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

Data Availability Statement

Data is available at https://doi.org/10.6084/m9.figshare.28807286.v1, .

Supplemental Material

Supplemental material for this article is available online.

Note

Anastasiia Galaktionova joined the University College Dublin (UCD) Spatial Dynamics Lab, School of Architecture, Planning and Environmental Policy, as a PhD researcher in 2023. Her research is centred on spatial analysis of street vitality and user experience, delving into how data-driven user experiences, perceptions, and attitudes can be integrated into urban planning.

Aura-Luciana Istrate is an Assistant Professor in Sustainable Urbanism and Urban Planning at UCD. She is the Primary Coordinator of C-NEWTRAL Doctoral Network and co-leading the Horizon Europe project REALLOCATE. Aura is leading the Active Research Group within Spatial Dynamics Lab at UCD, with resarch interests concerning street vitality, active mobility, nature-based solutions in built environments, community engagement, and overall improving urban liveability.

References

Appleyard

(2020) LIVABLE STREETS 2.0. Elsevier.

Austrian

Rosentraub

(2002) Cities, sports, and economic change: a retrospective assessment. Journal of Urban Affairs 24(5): 549–563.

Balsas

CJL

(2019) Walkable Cities: Revitalization, Vibrancy, and Sustainable Consumption. SUNY Press.

Baptista Neto

Barbosa

(2016) Impacts of traffic calming interventions on urban vitality. Proceedings of the Institution of Civil Engineers - Urban Design and Planning 169(2): 78–90.

Barreca

Curto

Rolando

(2020) Is the real estate market of new housing stock influenced by urban vibrancy? Complexity 2020: e1908698.

Delclòs-Alió

Gutiérrez

Miralles-Guasch

(2019) The urban vitality conditions of Jane Jacobs in Barcelona: residential and smartphone-based tracking measurements of the built environment in a Mediterranean metropolis. Cities 86: 220–228.

Eraydin

Özatağan

(2021) Pathways to a resilient future: a review of policy agendas and governance practices in shrinking cities. Cities 115: 103226.

Fang

Wang

(2021) Spatial characterization of urban vitality and the association with various street network metrics from the multi-scalar perspective. Frontiers in Public Health 9: 677910.

Frank

Sallis

Saelens

, et al. (2010) The development of a walkability index: application to the Neighborhood Quality of Life Study. British Journal of Sports Medicine 44(13): 924–933.

10.

Galaktionova

(2024) Street Vitality Functional Density Dataset. Figshare. [Data set]. DOI: 10.6084/m9.figshare.28807286.v1.

11.

Gehl

(2010) Cities for People. 1st edition. Island Press.

12.

Gómez-Varo

Delclòs-Alió

Miralles-Guasch

(2022) Jane Jacobs reloaded: a contemporary operationalization of urban vitality in a district in Barcelona. Cities 123: 103565.

13.

Graham

Ehlenz

Han

(2023) Professional sports venues as catalysts for revitalization? Perspectives from industry experts. Journal of Urban Affairs 45(10): 1841–1859.

14.

Grinberger

Minghini

Juhász

, et al. (2022) OSM science—the academic study of the OpenStreetMap project, data, contributors, community, and applications. ISPRS International Journal of Geo-Information 11(4): 230.

15.

Han

Qin

Xiao

, et al. (2023) The nonlinear relationships between built environment features and urban street vitality: a data-driven exploration. Environment and Planning B: Urban Analytics and City Science 51: 23998083231172985.

16.

Hao

Zhu

Zhong

(2015) The rise of big data on urban studies and planning practices in China: review and open research issues. Journal of Urban Management 4(2): 92–124.

17.

Harvey

Aultman-Hall

(2016) Measuring urban streetscapes for livability: a review of approaches. The Professional Geographer 68(1): 149–158.

18.

Herfort

Lautenbach

Porto de Albuquerque

, et al. (2023) A spatio-temporal analysis investigating completeness and inequalities of global urban building data in OpenStreetMap. Nature Communications 14(1): 3985.

19.

Huang

Wang

, et al. (2023) How diversity and accessibility affect street vitality in historic districts? Land 12(1): 219.

20.

Istrate

Chen

(2022) Liveable streets in Shanghai: Definition, characteristics and design. Progress in Planning 158: 100544.

21.

Istrate

Chen

Kadetz

, et al. (2021) Developing an analytical framework for liveable streets in Shanghai. Urban Design International 26(1): 3–20.

22.

Jacobs

(1961) The Death and Life of Great American Cities. New York: Random House.

23.

Jiang

Han

Liu

, et al. (2022) Street vitality and built environment features: a data-informed approach from fourteen Chinese cities. Sustainable Cities and Society 79: 103724.

24.

Klinmalai

Kaewlai

(2023) Major event-led strategy for urban development and revitalization. Journal of Architectural/Planning Research and Studies 20(2): 123–142.

25.

Kyttä

Broberg

Haybatollahi

, et al. (2016) Urban happiness: context-sensitive study of the social sustainability of urban settings. Environment and Planning B: Planning and Design 43(1): 34–57.

26.

Larimian

Sadeghi

(2021) Measuring urban social sustainability: scale development and validation. Environment and Planning B: Urban Analytics and City Science 48(4): 621–637.

27.

Pan

(2023) Assessment of influence mechanisms of built environment on street vitality using multisource spatial data: a case study in Qingdao, China. Sustainability 15(2): 1518.

28.

Long

Huang

(2019) Does block size matter? The impact of urban design on economic vitality for Chinese cities. Environment and Planning B: Urban Analytics and City Science 46(3): 406–422.

29.

Huang

Shi

, et al. (2019a) Exploring the associations between urban form and neighborhood vibrancy: a case study of Chengdu, China. ISPRS International Journal of Geo-Information 8(4): 165.

30.

Shi

Yang

(2019b) Impacts of built environment on urban vitality: regression analyses of Beijing and Chengdu, China. International Journal of Environmental Research and Public Health 16(23): 4592.

31.

Mehta

(2007) Lively streets: determining environmental characteristics to support social behavior. Journal of Planning Education and Research 27(2): 165–187.

32.

Montgomery

(1998) Making a city: urbanity, vitality and urban design. Journal of Urban Design 3(1): 93–116.

33.

Niu

, et al. (2022) The structural dimensions and community vibrancy: an exploratory analysis in Guangzhou, China. Cities 127: 103771.

34.

Oliver

Schuurman

Hall

(2007) Comparing circular and network buffers to examine the influence of land use on walking for leisure and errands. International Journal of Health Geographics 6(1): 41.

35.

Park

S-H

Kim

J-H

Choi

Y-M

, et al. (2013) Design elements to improve pleasantness, vitality, safety, and complexity of the pedestrian environment: evidence from a Korean neighbourhood walkability case study. International Journal on the Unity of the Sciences 17(1): 142–160.

36.

Priemus

Konings

(2000) Public transport in urbanised regions: the missing pursuit of the economic of cities link in the vitality. Planning Practice and Research. Epub ahead of print 1 August 2000. DOI: 10.1080/713691899.

37.

Priemus

Konings

(2001) Light rail in urban regions: what Dutch policymakers could learn from experiences in France, Germany and Japan. Journal of Transport Geography 9(3): 187–198.

38.

Radziszewska-Zielina

Adamkiewicz

Szewczyk

, et al. (2022) Decision-making support for housing projects in post-industrial areas. Sustainability 14(6): 3573.

39.

Ratcliffe

Flanagan

(2004) Enhancing the vitality and viability of town and city centres: the concept of the business improvement district in the context of tourism enterprise. Property Management 22(5): 377–395.

40.

Ravenscroft

(2000) The vitality and viability of town centres. Urban Studies 37(13): 2533–2549.

41.

Rey

Anselin

(2007) PySAL: a python library of spatial analytical methods. The Review of Regional Studies 37: 5–27.

42.

Sanders

Zuidgeest

Geurs

(2015) Liveable streets in Hanoi: a principal component analysis. Habitat International 49: 547–558.

43.

Shiftan

Burd-Eden

(2001) Modeling response to parking policy. Transportation Research Record: Journal of the Transportation Research Board 1765(1765): 27–34.

44.

Sipos-Onyestyák

Szikora

Szekeres

(2024) Sports facility development and politics in Budapest since 1945. International Journal of the History of Sport 1: 1–24.

45.

Still

Simmonds

(2000) Parking restraint policy and urban vitality. Transport Reviews 20(3): 291–316.

46.

Sung

H-G

Lee

(2015) Residential built environment and walking activity: empirical evidence of Jane Jacobs’ urban vitality. Transportation Research Part D: Transport and Environment 41: 318–329.

47.

Tang

Lin

, et al. (2018) Exploring the influence of urban form on urban vibrancy in Shenzhen based on mobile phone data. Sustainability 10(12): 4565.

48.

Tomalin

(1997) Town centre health checks:-some developments from practice. Planning Practice and Research 12(4): 383–392.

49.

Zhu

Xia

, et al. (2020) Portraying the spatial dynamics of urban vibrancy using multisource urban big data. Computers, Environment and Urban Systems 80: 101428.

50.

Vukmirović

Nikolić

(2023) Industrial heritage preservation and the urban revitalisation process in Belgrade. Journal of Urban Affairs 45(2): 191–216.

51.

Wang

Wong

Goh

(2021) Perceived importance of inclusive street dimensions: a public questionnaire survey from a vision(ing) perspective. Transportation 48(2): 699–721.

52.

Wang

Lei

Xue

, et al. (2022) Elaborating spatiotemporal associations between the built environment and urban vibrancy: a case of Guangzhou City, China. Chinese Geographical Science 32(3): 480–492.

53.

Wang

Liu

, et al. (2024) Identification of 71 factors influencing urban vitality and examination of their spatial dependence: a comprehensive validation applying multiple machine-learning models. Sustainable Cities and Society 108: 105491.

54.

Wangbao

(2022) Spatial impact of the built environment on street vitality: a case study of the Tianhe District, Guangzhou. Frontiers in Environmental Science 10: 966562.

55.

Xia

Zhang

, et al. (2022) The spatial pattern and influence mechanism of urban vitality: a case study of Changsha, China. Frontiers in Environmental Science 10: 942577.

56.

Yang

Jiao

(2021) Exploring the correlation between block vitality and block environment based on multisource big data: taking Wuhan City as an Example. Land 10(9): 984.

57.

Yang

Wang

Qin

, et al. (2022) Analysis of urban vitality in Nanjing based on a plot boundary-based neural network weighted regression model. ISPRS International Journal of Geo-Information 11(12): 624.

58.

Cui

, et al. (2022) TOD and vibrancy: the spatio-temporal impacts of the built environment on vibrancy. Frontiers in Environmental Science 10.

59.

Yuan

Chen

(2021) Analysis of factors influencing street vitality in high-density residential areas based on multi-source data: a case study of Shanghai. International Journal of High-Rise Buildings 10(1): 1–8.

60.

Yue

Zhu

(2019) Exploring the relationship between urban vitality and street centrality based on social network review data in Wuhan, China. Sustainability 11(16): 4356.

61.

Yue

Zhuang

Yeh

AGO

, et al. (2017) Measurements of POI-based mixed use and their relationships with neighbourhood vibrancy. International Journal of Geographical Information Science 31(4): 658–675.

62.

Zacharias

(2020) The contribution of a tramway to pedestrian vitality. TeMA - Journal of Land Use, Mobility and Environment 13(3): 445–457.

63.

Zacharopoulou

Skopeliti

Nakos

(2021) Assessment and visualization of OSM consistency for European cities. ISPRS International Journal of Geo-Information 10(6): 361.

64.

Zhang

Sun

Chan

, et al. (2021) Exploring impact of surrounding service facilities on urban vibrancy using Tencent location-aware data: a case of Guangzhou. Sustainability 13(2): 444.

65.

Zhang

Zhou

Brovelli

, et al. (2022) Assessing OSM building completeness using population data. International Journal of Geographical Information Science 36(7): 1443–1466.

66.

Zheng

Tang

(2023) The vitality of public space and the effects of environmental factors in Chinese suburban rural communities based on tourists and residents. International Journal of Environmental Research and Public Health 20(1): 263.

67.

Zhu

Liu

, et al. (2021) Smart city oriented optimization of residential blocks on intensive urban sensing data based on fuzzy evaluation algorithm. Journal of Ambient Intelligence and Humanized Computing 12(1): 911–921.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

1.84 MB