Introduction
Bus lanes have the potential to improve the reliability of bus operations by limiting the negative impacts of traffic congestion, reducing the variability of travel times, and increasing the average commercial speed - which can ultimately be used to increase service frequency.
However, introducing them on road infrastructure affects other modes, such as private vehicles, which may experience a decrease in the level of service due to reduced allocated space, potentially jeopardizing public acceptance.
Common criteria for implementing bus lanes include:
- Frequency of buses (and trams) per hour and direction, at a peak hour;
- Number of lanes in the same direction;
- Existing traffic conditions;
- Existing bus lanes in the area (from a network continuity perspective).
GTFShift provides methods to analyse these dimensions, namely:
GTFShift::get_way_frequency_hourly(), to obtain the frequency of services per hour for each road segment with transit service and its associated characteristics (such as number of lanes).GTFShift::osm_bus_lanes(), to identify existing bus lanes in the road network.GTFShift::rt_collect_json()orGTFShift::rt_collect_protobuf(), to collect GTFS-RT data, which can be later used withGTFShift::rt_extend_prioritization()to include real-time operational metrics in the prioritization analysis.
This document explores how to use these methods in a combined way to assist public transport planners in prioritizing bus lane implementations. For details on the several encapsulated features and method variations, refer to the numbered articles in the menu, that explore in detail each of the specific approaches followed.
Prioritize lanes
Generate base indicators
GTFShift::prioritize_lanes() is a simple method that
generates indicators for most of the criteria mentioned above using GTFS
and OpenStreetMaps data (service frequency and lane characteristics).
With a single call, it returns a data.frame with the relevant metrics
for each road segment with transit service.
# Get GTFS from library GTFS database for Portugal
data = read.csv(system.file("extdata", "gtfs_sources_pt.csv", package = "GTFShift"))
gtfs_id = "lisboa"
gtfs = GTFShift::load_feed(data$URL[data$ID == gtfs_id], create_transfers=FALSE)
osm_q = opq(bbox=sf::st_bbox(tidytransit::shapes_as_sf(gtfs$shapes))) |>
add_osm_feature(key = "route", value = c("bus", "tram")) |>
add_osm_feature(key = "network", value = "Carris", key_exact = TRUE)
lanes = prioritize_lanes(gtfs, osm_q)
summary(lanes)
#> way_osm_id hour frequency is_bus_lane
#> Length:132056 Min. : 0.00 Min. : 1.000 Mode :logical
#> Class :character 1st Qu.: 8.00 1st Qu.: 3.000 FALSE:120040
#> Mode :character Median :13.00 Median : 6.000 TRUE :12016
#> Mean :12.56 Mean : 9.673
#> 3rd Qu.:18.00 3rd Qu.:13.000
#> Max. :23.00 Max. :99.000
#> n_lanes_parking n_lanes_circulation n_lanes n_directions
#> Min. :0.00000 Min. :1.00 Min. :1.000 Min. :1.000
#> 1st Qu.:0.00000 1st Qu.:1.00 1st Qu.:1.000 1st Qu.:1.000
#> Median :0.00000 Median :2.00 Median :2.000 Median :1.000
#> Mean :0.01438 Mean :2.15 Mean :2.165 Mean :1.364
#> 3rd Qu.:0.00000 3rd Qu.:3.00 3rd Qu.:3.000 3rd Qu.:2.000
#> Max. :2.00000 Max. :7.00 Max. :7.000 Max. :2.000
#> n_lanes_circulation_direction n_lanes_direction routes
#> Min. :1.000 Min. :1.000 Length:132056
#> 1st Qu.:1.000 1st Qu.:1.000 Class :character
#> Median :1.000 Median :1.000 Mode :character
#> Mean :1.703 Mean :1.713
#> 3rd Qu.:2.000 3rd Qu.:2.000
#> Max. :6.000 Max. :6.000
#> geom
#> LINESTRING :132056
#> epsg:4326 : 0
#> +proj=long...: 0
#>
#>
#> Extend with GTFS-RT data
If GTFS-RT data is available, it can be used to extend the prioritization analysis with real-time operational metrics, such as average speed. This can help identify road segments where buses are experiencing significant delays due to traffic congestion, which may benefit from bus lane implementation.
rt_collection = read.csv("rt_collect_file.csv") |>
sf::st_as_sf(coords = c("vehicle.position.longitude", "vehicle.position.latitude"), crs = 4326)
lanes = GTFShift::rt_extend_prioritization(
lane_prioritization = lanes,
rt_collection = rt_collection
)Refer to the GTFS Real Time article for details on how to collect GTFS-RT data and extend the prioritization analysis.
Visualize results
The aggregated data can then be manipulated according to the prioritization criteria defined by the user. For instance, the following code highlights (in red) the road segments as high priority for bus lane implementation if they have more than 1 lane per direction and a frequency above the median number of buses per hour registered at 8:00.
lanes_0800 = lanes |> filter(hour==8)
summary(lanes_0800)
#> way_osm_id hour frequency is_bus_lane
#> Length:6665 Min. :8 Min. : 1.00 Mode :logical
#> Class :character 1st Qu.:8 1st Qu.: 5.00 FALSE:6114
#> Mode :character Median :8 Median :10.00 TRUE :551
#> Mean :8 Mean :13.32
#> 3rd Qu.:8 3rd Qu.:18.00
#> Max. :8 Max. :99.00
#>
#> n_lanes_parking n_lanes_circulation n_lanes n_directions
#> Min. :0.00000 Min. :1.000 Min. :1.000 Min. :1.000
#> 1st Qu.:0.00000 1st Qu.:1.000 1st Qu.:1.000 1st Qu.:1.000
#> Median :0.00000 Median :2.000 Median :2.000 Median :1.000
#> Mean :0.01425 Mean :2.122 Mean :2.136 Mean :1.365
#> 3rd Qu.:0.00000 3rd Qu.:3.000 3rd Qu.:3.000 3rd Qu.:2.000
#> Max. :2.00000 Max. :7.000 Max. :7.000 Max. :2.000
#>
#> n_lanes_circulation_direction n_lanes_direction routes
#> Min. :1.000 Min. :1.000 Length:6665
#> 1st Qu.:1.000 1st Qu.:1.000 Class :character
#> Median :1.000 Median :1.000 Mode :character
#> Mean :1.679 Mean :1.689
#> 3rd Qu.:2.000 3rd Qu.:2.000
#> Max. :6.000 Max. :6.000
#>
#> speed_avg speed_median speed_p25 speed_p75
#> Min. : 0.01694 Min. : 0.000 Min. : 0.000 Min. : 0.01694
#> 1st Qu.: 8.53002 1st Qu.: 7.994 1st Qu.: 6.138 1st Qu.:10.19889
#> Median : 9.83646 Median : 9.259 Median : 7.251 Median :11.79780
#> Mean :10.24983 Mean : 9.589 Mean : 7.377 Mean :12.39146
#> 3rd Qu.:11.48475 3rd Qu.:10.886 3rd Qu.: 8.469 3rd Qu.:13.90043
#> Max. :64.79895 Max. :44.118 Max. :41.508 Max. :77.74918
#> NA's :128 NA's :128 NA's :128 NA's :128
#> speed_count route_names geom
#> Min. : 1.0 Length:6665 LINESTRING :6665
#> 1st Qu.: 42.0 Class :character epsg:4326 : 0
#> Median : 101.0 Mode :character +proj=long...: 0
#> Mean : 176.5
#> 3rd Qu.: 217.0
#> Max. :5541.0
#> NA's :128
p50_frequency = quantile(lanes_0800$frequency, 0.5, na.rm=TRUE)
p50_speed = quantile(lanes_0800$speed_avg, 0.5, na.rm=TRUE)
mapview::mapview(
lanes_0800 |> filter(is_bus_lane & (frequency<p50_frequency | (is.na(n_lanes) | n_lanes_direction<=1) | speed_avg<=p50_speed)),
layer.name=sprintf("Bus lane with -%d bus/h OR -2 lane/dir OR %.2f km/h or - avg. speed", p50_frequency, p50_speed),
color="#DAD887",
homebutton=FALSE,
lwd=3
) + mapview::mapview(
lanes_0800 |> filter(is_bus_lane & frequency>=p50_frequency & !is.na(n_lanes) & n_lanes_direction>1 & speed_avg>p50_speed),
layer.name=sprintf("Bus lane with +%d bus/h AND +1 lane/dir AND +%.2f km/h avg.speed", p50_frequency-1, p50_speed),
color="#3BC1A8",
homebutton=FALSE,
lwd=3
) + mapview::mapview(
lanes_0800 |> filter(!is_bus_lane & frequency>=p50_frequency & !is.na(n_lanes) & n_lanes_direction>1 & speed_avg<=p50_speed),
layer.name=sprintf("NO bus lane with +%d bus/h AND +1 lane/dir AND %.2f km/h or - avg.speed", p50_frequency-1, p50_speed),
color="#F63049",
homebutton=FALSE,
lwd=3
)This visual representation allows to easily identify not only the high-priority segments for bus lane implementation, but also their spatial distribution across the existent network. A process that extends the results by incorporating the network continuity perspective, enabling the identification and eventual prioritization of critical segments that connect bus lanes but have a bad performance.