trafficdb

The Traffic Database aims to bring GTFS, OSM and Traffic (Speed) data together. This project grew out of the US-FTA's Integrated Mobility Innovation (IMI) program.

Example stop & traffic maps:

• TriMet (Portland, OR)
• C-Tran (Vancouver, WA)
• RVTD (Medford, OR)
• ACTransit (Oakland, CA)
• VTA (San Jose, CA)
• RTD (Denver, CO)
• COTA (Columbus, OH)

Manual Install

• prerequisite install: git, postgres, psql, postgis, ogr2ogr (gdal), python 3.x, psycopg2-binary, zc.buildout.
  • note: there is a docker-compose alternative to installing PostGIS, gdal and ogr2ogr (see below)
  • for python, it should be just a matter of getting python installed, then doing a pip install psycopg2-binary zc.buildout
• clone the trafficdb repo: git clone https://github.com/OpenTransitTools/trafficdb.git
• cd trafficdb
• run the scripts/db/create.sh shell script
• run buildout to pull in python dependencies
• add your INRIX credentials to trafficdb/config/base.ini
  • [inrix]
  • vendorid = ___
  • consumerid = ___
  • note: if you don't do this, you won't see any speed data
  • git update-index --assume-unchanged config/base.ini
• run bin/load_all -c -g -s test -d postgres://ott@localhost:5432/ott ott/trafficdb/model/inrix/test/gtfs.zip
• run bin/load_speed_data -s test -d postgres://ott@localhost:5432/ott # note: run this every N minutes to get up-to-date speed data into the database

Docker-Compose alternative install of PostGIS and GDAL:

• install Docker and Docker-Compose
• cd transitdb
• docker-compose up
• . ./scripts/docker_dot_source_me.sh

Manual Install (real transit data load):

if the above load_all worked well, then the next step will be to load a full dataset

• grab INRIX's .geojson traffic segment data from https://map-data-downloader.inrix.com/
  • for the example below, using TriMet's latest GTFS file, you should download and unzip the USA_Oregon_geojson.zip file

https://developer.trimet.org/GTFS.shtml bin/load_all -c -g -s trimet -t $PWD/USA_Oregon.geojson -d postgres://ott@localhost:5432/ott https://developer.trimet.org/schedule/gtfs.zip bin/load_speed_data -s trimet

https://www.rvtd.org/Page.asp?NavID=60 bin/load_all -c -g -s rvtd -t $PWD/USA_Oregon.geojson -d postgres://ott@localhost:5432/ott http://feed.rvtd.org/googleFeeds/static/google_transit.zip bin/load_speed_data -s rvtd

https://www.c-tran.com/about-c-tran/business/c-tran-gtfs-data bin/load_all -c -g -s ctran -t $PWD/USA_Washington.geojson $PWD/USA_Oregon.geojson -d postgres://ott@localhost:5432/ott https://www.c-tran.com/images/Google/GoogleTransitUpload.zip bin/load_speed_data -s ctran

https://gtfs.vta.org/ bin/load_all -c -g -s vta -t $PWD/USA_CA_BayArea.geojson -d postgres://ott@localhost:5432/ott https://gtfs.vta.org/gtfs_vta.zip bin/load_speed_data -s vta

http://www.actransit.org/planning-focus/data-resource-center/ bin/load_all -c -g -s actransit -t $PWD/USA_CA_BayArea.geojson -d postgres://ott@localhost:5432/ott https://url.actransit.org/GtfsCurrent bin/load_speed_data -s actransit

https://www.rtd-denver.com/business-center/open-data/gtfs-developer-guide curl https://www.rtd-denver.com/files/gtfs/google_transit.zip > x # note: gtfsdb's current HTTP GET routine doesn't work with RTD's server / redirects bin/load_all -c -g -s rtd -t $PWD/USA_Colorado.geojson -d postgres://ott@localhost:5432/ott x bin/load_speed_data -s rtd

https://www.cota.com/data/ bin/load_all -c -g -s cota -t $PWD/USA_Ohio.geojson -d postgres://ott@localhost:5432/ott https://www.cota.com/COTA/media/COTAContent/OpenGTFSData.zip bin/load_speed_data -s cota

Individual Processes: CLEANUP TODO

• step 1: load gtfs data ... calculate all stop-to-stop segments in the data
  • bin/load_gtfs_data -c -g -s trimet -d postgres://ott@localhost/ott https://developer.trimet.org/schedule/gtfs.zip
• step 2: load traffic vendor street / segment data (INRIX in this case)
  • ott/trafficdb/model/inrix/load_inrix_geojson.sh INRIX/USA_*.geojson
• step 3: match/conflate stop-segments with traffic segment data
  • bin/match_segments
• step 4: load speed data from INRIX into db (on-going ... run every 5 mintutes?)
  • bin/load_speed_data
• step 5: plot lastest segment / speed data on the example map
  • bin/generate-speed-geojson --show-map

note: the load will take upwards of 1 hour (or more) ... but once done, you'll have a complete database with transit and speed data

todo:

• speed and osm data (example datasets) needed to complete things
• create a docker container for the steps above!
• create a second & third example around MapBox and/or TomTom data