Using OpenStreetMap data to assess the bikeability of schools

Harald Kliems
kliems@wisc.edu

UW–Madison Prevention Research Center

UW–Madison School of Medicine and Public Health, Dept. of Pediatrics

Ben Varick
ben@wisconsinbikefed.org

Wisconsin Bike Fed

UW–Madison Nelson Institute for Environmental Studies

Samuel Younkin
syounkin@wisc.edu

UW–Madison Nelson Institute for Environmental Studies

Our research question

  • Biking to school has many benefits
  • Lack of safe routes is a major barrier (for parents and kids)
  • We can use OSM data to develop quantitative measures of the bikeability of schools!

School locations and bikeable area

  • School locations and the area that is within a 3 mile bike ride to the school
  • Attendance zones - relevant for some school districts

Census data and grid cells

  • Student population from American Community Survey
  • Uses bike and attendance areas to “assign” cells to schools

Zooming in

Routing with BRouter

  • Routing with BRouter
  • Calculate routes with two profiles: safety & shortest

Traffic stress with BrokenSpoke Analyzer

  • Bike Level of Traffic Stress classification w/ BrokenSpoke Analyzer
  • Classifies segments and intersections into high stress or low stress
  • Based on number of lanes, posted speed limit, type of bike lanes, signals, etc.
        -- conventional bike lane
        WHEN ft_bike_infra = 'lane'
            THEN CASE
                WHEN COALESCE(speed_limit, :default_speed) > 20 THEN 3
                WHEN COALESCE(speed_limit, :default_speed) <= 20
                    THEN CASE
                        WHEN COALESCE(ft_lanes, :default_lanes) > 1 THEN 3
                        ELSE CASE -- lanes = 1
                            WHEN
                                COALESCE(
                                    ft_bike_infra_width, :default_facility_width
                                )
                                >= 4 THEN 1
                            ELSE 3 -- less than 4 ft
                        END
                    END
                ELSE 3
            END
  • Brokenspoke Analyzer uses OSM tags to calculate a bicycle level of stress value
  • Example: highway=tertiary and higher with a bike lane:
    • Speed limit 20 mph or lower: low stress (1)
      • unless there’s more than one lane
      • unless there’s one lane but the bike lane is narrower than 4 ft

Aggregate routes

  • Prioritizing limited resources to best help kids bike to school.
  • Which streets are the biggest barriers to the most students?

Aggregate across schools & district

  • In which neighborhoods is biking to school the most dangerous?
  • Can guide building new schools and housing density

Aggregate across schools & district

  • Prioritize education and encouragement efforts.
  • Which schools have high bike potential?
  • How does bikability relate to other social characteristics?

What-if analysis

  • Routing is done locally using an .osm.pbf file, editable!
  • Using osmChange files, looking for better workflow
  • How many students will be impacted by a change?
  • Focus groups with high school students to identify barriers

A photograph of a street with many cars and painted bike lanes

Aberg Avenue, @shantalb - CC BY-SA - Mapillary

Conclusions and next steps

  • WE NEED ACTUAL DATA ON BIKING TO SCHOOL RATES
  • Dealing with mismatch between BrokeSpoke Analyzer Traffic Stress and BRouter safety profile
  • Take measures of OSM data quality into account
  • Make our tool available widely (and easy to use)
  • Our hope is this tool is useful for city staff, policy makers, advocates

Acknowledgments

This research was funded through a pilot grant from the Center for Health, Energy, and Environmental Research (CHEER; NIH grant #1P20ES036747-01). We thank all the high school students involved in the project for their contributions!

  • Talk to us!

  • Reach out:

    • kliems@wisc.edu
    • ben@wisconsinbikefed.org
    • syounkin@wisc.edu