Using bicycle app data to develop Safety Performance Functions (SPFs) for bicyclists at intersections: A generic framework

Abstract More accurate predictions of bicycle crashes can increase the return on investment from bicycle safety initiatives. One useful tool to understand the association between critical factors and crashes is Safety Performance Functions (SPFs), but most U.S. studies have developed SPFs for motorized vehicles not for bicycles. The objective of this study is to develop SPFs for intersections in a medium- and a large-sized city using bicycle app data (i.e., Strava data), leveraging the rising popularity of social media and mobile phones. Our case studies are from the Portland and Eugene-Springfield metropolitan areas with sizable bike population, which alleviates the challenge of insufficient bicycle volume and crash data. Specifically, in this research (1) bicycle SPFs are created for intersections in medium- and large-sized cities; (2) affordable bicycle app volume data, is used as a surrogate for exposure; (3) bicycle app data is shown to be correlated with automatic bike count station data ( p 0.01 ); (4) bicycle app count shows a significant association with intersection crashes ( p 0.01 ); (5) a general framework for building bicycle SPFs is developed for jurisdictions and the corresponding model is validated to demonstrate predictive ability; and (6) recommendations on infrastructure design and non-motorized policy making are provided with the goal of developing a safer environment for bicyclists.

[1]  David S. Hurwitz,et al.  Towards Effective Design Treatment for Right Turns at Intersections with Bicycle Traffic , 2015 .

[2]  Feng Wei,et al.  An empirical tool to evaluate the safety of cyclists: Community based, macro-level collision prediction models using negative binomial regression. , 2013, Accident; analysis and prevention.

[3]  T. Nelson,et al.  Correcting Bias in Crowdsourced Data to Map Bicycle Ridership of All Bicyclists , 2019, Urban Science.

[4]  Nataliya V Malyshkina,et al.  Zero-state Markov switching count-data models: an empirical assessment. , 2008, Accident; analysis and prevention.

[5]  F Mannering,et al.  Modeling accident frequencies as zero-altered probability processes: an empirical inquiry. , 1997, Accident; analysis and prevention.

[6]  Q. Vuong.  Likelihood Ratio Tests for Model Selection and Non-Nested Hypotheses , 1989 .

[7]  Krista Nordback,et al.  Bicyclist safety performance functions for a U.S. city. , 2014, Accident; analysis and prevention.

[8]  H. Akaike.  A new look at the statistical model identification , 1974 .

[9]  Yiyi Wang,et al.  Does bicycle network level of traffic stress (LTS) explain bicycle travel behavior? Mixed results from an Oregon case study , 2016 .

[10]  Luis F Miranda-Moreno,et al.  Mapping cyclist activity and injury risk in a network combining smartphone GPS data and bicycle counts. , 2015, Accident; analysis and prevention.

[11]  Kari Watkins,et al.  Comparison of Smartphone-based Cyclist GPS Data Sources , 2016 .

[12]  Myungseob Kim,et al.  Assessing Critical Factors Associated with Bicycle Collisions at Urban Signalized Intersections , 2008 .

[13]  Anthony C. Davison,et al.  Bootstrap Methods and Their Application , 1998 .

[14]  Erick Guerra,et al.  Where do bike lanes work best? A Bayesian spatial model of bicycle lanes and bicycle crashes. , 2017, Safety science.

[15]  J. Mullahy.  Specification and testing of some modified count data models , 1986 .

[16]  Peng Chen,et al.  Built environment effects on cyclist injury severity in automobile-involved bicycle crashes. , 2016, Accident; analysis and prevention.

[17]  Jennifer Dill,et al.  Where do cyclists ride? A route choice model developed with revealed preference GPS data , 2012 .

[18]  Greg Lindsey,et al.  Urban trails and demand response to weather variations , 2018, Transportation Research Part D: Transport and Environment.

[19]  Chen Chen,et al.  Development of a Crash Risk-Scoring Tool for Pedestrian and Bicycle Projects in Oregon , 2018 .

[20]  Shanthi Ameratunga,et al.  Incidence, risk, and protective factors of bicycle crashes: findings from a prospective cohort study in New Zealand. , 2013, Preventive medicine.

[21]  Meghan Winters,et al.  Comparing the effects of infrastructure on bicycling injury at intersections and non-intersections using a case–crossover design , 2013, Injury Prevention.

[22]  Lars Ekman,et al.  ON THE TREATMENT OF FLOW IN TRAFFIC SAFETY ANALYSIS: A NON-PARAMETRIC APPROACH APPLIED ON VULNERABLE ROAD USERS , 1996 .

[23]  Walter Musakwa,et al.  THE POTENTIAL OF STRAVA DATA TO CONTRIBUTE IN NON-MOTORISED TRANSPORT (NMT) PLANNING IN JOHANNESBURG , 2016 .

[24]  Wesley E. Marshall,et al.  Estimating Annual Average Daily Bicyclists , 2013 .

[25]  D. L. Robinson.  Safety in numbers in Australia: more walkers and bicyclists, safer walking and bicycling. , 2005, Health promotion journal of Australia : official journal of Australian Association of Health Promotion Professionals.

[26]  Chen Chen,et al.  How bicycle level of traffic stress correlate with reported cyclist accidents injury severities: A geospatial and mixed logit analysis. , 2017, Accident; analysis and prevention.

[27]  Michelle DeRobertis,et al.  The safety of urban cycle tracks: a review of the literature. , 2013, Accident; analysis and prevention.

[28]  Rune Elvik,et al.  Safety-in-numbers: A systematic review and meta-analysis of evidence , 2017 .

[29]  Randall Guensler,et al.  Safety impacts of bicycle infrastructure: A critical review. , 2017, Journal of safety research.

[30]  Thomas Jonsson,et al.  Development of a Safety Performance Function for Motorcycle Accident Fatalities on Malaysian Primary Roads , 2013 .

[31]  Maria Börjesson,et al.  The value of time and external benefits in bicycle appraisal , 2012 .

[32]  Jeffrey M Casello,et al.  Modeling Cyclists’ Route Choice Based on GPS Data , 2014 .

[33]  Trisalyn A. Nelson,et al.  Mapping ridership using crowdsourced cycling data , 2016 .

[34]  K. Sælensminde.  COST-BENEFIT ANALYSES OF WALKING AND CYCLING TRACK NETWORKS TAKING INTO ACCOUNT INSECURITY, HEALTH EFFECTS AND EXTERNAL COSTS OF MOTORIZED TRAFFIC , 2004 .

[35]  P. Burman.  A comparative study of ordinary cross-validation, v-fold cross-validation and the repeated learning-testing methods , 1989 .

[36]  William W. Hunter,et al.  Police Reporting of Pedestrians and Bicyclists Treated in Hospital Emergency Rooms , 1998 .

[37]  R. R. Hocking.  The analysis and selection of variables in linear regression , 1976 .

[38]  Bo Lan,et al.  In Pursuit of Safety: Systemic Bicycle Crash Analysis in Seattle, WA , 2017 .

[39]  J. Hilbe.  Negative Binomial Regression: Preface , 2007 .

[40]  Tarek Sayed,et al.  Safety performance functions using traffic conflicts , 2013 .

[41]  Billy Charlton,et al.  A GPS-based bicycle route choice model for San Francisco, California , 2011 .

[42]  J. Bared,et al.  Accident Models for Two-Lane Rural Segments and Intersections , 1998 .

[43]  S. Miaou.  The relationship between truck accidents and geometric design of road sections: Poisson versus negative binomial regressions. , 1994, Accident; analysis and prevention.

[44]  R. Dicker,et al.  Using trauma center data to identify missed bicycle injuries and their associated costs , 2012, The journal of trauma and acute care surgery.

[45]  Baoshan Huang,et al.  Multivariate random-parameters zero-inflated negative binomial regression model: an application to estimate crash frequencies at intersections. , 2014, Accident; analysis and prevention.

[46]  Graham R. Wood,et al.  Safety Performance Functions for Bicycle Crashes in New Zealand and Australia , 2011 .

[47]  S. Washington,et al.  Statistical and Econometric Methods for Transportation Data Analysis , 2010 .