Pavement widths

[mvl22]

Given the current concern about physical distancing on pavements, it would be good if we can get the calculated pavement width data public. This presumably requires OS clearance.

The issue of single/both sides of the street would need to be dealt with, including dual carriageways.

See also #24.

See also https://twitter.com/AsEasyAsRiding/status/1252647649997701120 referring to http://www.sidewalkwidths.nyc/ which is documented at: https://github.com/meliharvey/sidewalkwidths-nyc/blob/master/README.md

[metazool]

https://twitter.com/OrdnanceSurvey/status/1253611897267597314 FWIW the OS has a covid-19 response specific licence which they have stated would cover the data in this project.

[Robinlovelace]

Many thanks for the follow-up @metazool !

[Robinlovelace]

@mvl22, @metazool and @mem48 could we use the road width data plus overlay of smoothed cycling potential estimates to provide an evidence-base for identifying places where temp. cycleways could go in? Basically roads with 2 carriageways and with high cycling potential like Scott Hall Road.

[Robinlovelace]

Heads-up @mvl22 and @mem48 here is a reproducible demo showing how to find roads that are wide and those with more than 2 lanes, indicating at least one spare lane for temporary cycleways or increased pavement space:

# Aim: identify wide roads that have high cycling potential

# setup -------------------------------------------------------------------

library(tidyverse)
library(sf)
#> Linking to GEOS 3.8.0, GDAL 3.0.4, PROJ 7.0.0
library(tmap)
tmap_mode("view")
#> tmap mode set to interactive viewing

# input data --------------------------------------------------------------

up = "https://github.com/cyipt/cyipt-bigdata/raw/master/osm-prep/Leeds/osm-lines.Rds"
roads = readRDS(url(up))
names(roads)
#>  [1] "osm_id"             "name"               "ref"               
#>  [4] "highway"            "junction"           "roadtype"          
#>  [7] "onewaysummary"      "elevation"          "maxspeed"          
#> [10] "segregated"         "sidewalk"           "cycleway.left"     
#> [13] "lanes.psv.forward"  "lanes.forward"      "lanes.backward"    
#> [16] "lanes.psv.backward" "cycleway.right"     "region"            
#> [19] "quietness"          "aadt"               "ncycles"           
#> [22] "id"                 "pct.census"         "pct.gov"           
#> [25] "pct.gen"            "pct.dutch"          "pct.ebike"         
#> [28] "pct.total"          "width"              "widthpath"         
#> [31] "ncollisions"        "bikeCas"            "totalCas"          
#> [34] "totalVeh"           "geometry"
summary(roads$width)
#>    Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
#>    0.60    5.30    6.60    6.74    8.00   28.60   25355
wide_potential1 = roads %>% filter(width > 10, pct.dutch > 100)
mapview::mapview(wide_potential1)

wide_potential2 = roads %>% filter(lanes.forward > 1, pct.dutch > 100)
mapview::mapview(wide_potential2)

wide_potential3 = roads %>% filter(lanes.forward + lanes.backward > 2, pct.dutch > 100)
mapview::mapview(wide_potential3)

^{Created on 2020-04-27 by the reprex package (v0.3.0)}

[Robinlovelace]

Roads with high cycling potential and more than 1 lane in major cities such as Leeds, Liverpool, London, Bristol, Birmingham, Manchester, Newcastle, Cardiff.

[mem48]

@Robinlovelace have you put in a formal application for a covid19 licence? If so lets get the latest mastermap and rerun the analysis. Onve the next release is done.

[Robinlovelace]

HI @mem48 yes I have requested the latest MM Topo layer. It was supposed to be 'delivered' on Friday via text. I'll check. They seem very happy for us to use it which is great.