SiteSelection

An open-source tool to identify complex areas for street space reallocation

Rosa Félix, Gabriel Valença, Filipe Moura, Ana Morais de Sá

Complex areas

• Space is limited, and there is competition 🚶‍♂️🚗🚲🚌🎍

• Growing need for more sustainable, efficient, and equitable space distribution

• Challenge of reallocating street space, dynamically over time

• Complexity of identifying areas where intervention makes sense (Valença, Moura, and Sá 2024)

Combination of several types of data

A process to select the cell locations that meet a set o criteria for a given city or neighbourhood where the street space is more disputed.

• Population density (census)
• POIs and activities
• Public Transit Frequency (GTFS for bus and tram)
• Road network centrality measures
  • betweenness, closeness, and degree

Problem

Using point and click software to fetch and process data…

This is fine for a single case study 😀

But what it you want to replicate the exactly same process for 400 other locations? 🤯

SiteSelection

SiteSelection aims to identify complex areas suitable for street space reallocation.

GIS tools: methods to process existing geo-data and classify areas based on key metrics like road network centrality, population density, and public transport frequency.

Data Sources: open data such as OpenStreetMap road network and POIs, Census data, GTFS for public transport, administrative areas

Open-source tool: an R package 📦

Processes

Map

Returns a ranked list of complex areas based on their need for space reallocation (0 - no complexity, 6 - very complex)

Options

Thresholds for candidate selection

degree_min = mean           # mean or median? default: mean
betweeness_range = 0.40     # percentile to exclude (upper and lower) default: 0.25
closeness_range = 0.25      # percentile to exclude (upper and lower) default: 0.25
population_min = median     # mean or median? default: mean
entropy_min = 0.35          # value to exclude (lower) default: 0.5
freq_bus = c(4, 10, 20)     # frequency of bus stops to define level of service 

For different contexts, we can have different cut thresholds… Under assessment

Options

Grid types

• Squares
• Hexagons
• Universal h3 hexagon grid - useful to include other variables

Options

Work with non-administrative boundary

• Set another polygon limit to the analysis
• create a .geojson file with the area of interest

Options

Keep track of your analysis and export results

Live demo

• Modify a network parameter

• Modify the grid type or resolution

• Other case study

Conclusions

• Repeated processes and analysis, human error proof 🛡️
• Super fast 🚀
  • It would normally take days to run everything 😵
  • ~1min for a city
  • for 200 cities em Portugal < 45 min

Conclusions

• Change one parameter and re-run only what is needed
• Scalable and expandable, for other custom conditions and functions
• Replicable for other locations 🌎🌍🌏
  • We included scripts to prepare data (census, GTFS)
• Open source 📡

Future developments and applications

• Each function can standalone in the package

  • get_census()
  • get_osm(), clean_osm()
  • get_centrality()
  • make_grid()
  • get_transit()

• And be used for other purposes! out of the Streets4All scope

Thank you

More: u-shift.github.io/SiteSelection

contributions are welcome!

Rosa Félix - rosamfelix@tecnico.pt

About

This work is part of Streets4All Project, developed at the University of Lisbon and at the University of Coimbra, and funded by Fundação para a Ciência e Tecnologia (PTDC/ECI-TRA/3120/2021).

This work is part of the research activity carried out at Civil Engineering Research and Innovation for Sustainability (CERIS) and has been funded by Fundação para a Ciência e a Tecnologia (FCT), Portugal in the framework of project UIDB/04625/2020, project PTDC/ECI-TRA/3120/2021, and project 2022,07909,CEECIND/CP1713/CT0017.

The concept is based on:

Valença, Gabriel, Filipe Moura, and Ana Morais de Sá. 2024. “Where Is It Complex to Reallocate Road Space?” Environment and Planning B: Urban Analytics and City Science 51 (6): 1290–1307. https://doi.org/10.1177/23998083231217770.