Bicycle Route Choice Data – an overview of methodological approaches Ray Pritchard 14th September 2015 Cycling & Society Symposium, Manchester
Why is route choice important? • Assist cycling promotion • Existing data – static • Network impacts • Preferences of users: o Built environment o Aesthetics o Topography The$car$era$that$was$–$and$in$ some$areas,$s2ll$is.$ o Ease of navigability 2
Not all route choice studies are created equal • Hypothetical route choice (how would you … ) o Amongst real options o Amongst not yet implemented options (preference) • Actual route choice (observe/how did you … ) 3
Literature Review (19) Publica(on+ Year+ Authors+ Data+Region+ Technique+ Stated$preference$interview$?$door$to$door$to$persons$in$region$prior$to$bike$lane$installa2on.$N:$254$ 1978$ Lo:$and$Tardiff$ USA$?$Davis$ before,$110$aMer.$ 1984$ Bovy$and$Bradley$ NL?$DelM$ Stated$preference$survey.$N:$unknown$ 1986$ Axhausen$and$Smith$India$?$Mumbai$ Stated$preference$survey.$N:$unknown$ Paper?based$Surveys$distributed$via$mail,$in$person$and$in$bicycle$shops.$Guelph$Community$Bicycle$Survey$ and$Guelph$Bicycle$User$Survey.$N:$397$ 1997$ Aultman?Hall$ Canada$?$Guelph$ Canada$?$O:awa$ 1998$ Aultman?Hall$ &$Toronto$ Paper?based$Surveys.$Mail?back$ques2onnaire$a:ached$to$parked$bicycles.$N:2964$ 2005$ Raford$ UK$?$London$ Self?reported$revealed$preference$routes.$'Business$cycle'$project$of$employees$in$central$London.$N:423$ 2007$ Eisenman$ USA$?$Hanover$ Sensor$equipped$bicycle.$Demonstra2on$of$concept.$N:$5$bicycles$equipped.$ USA$?$ Data$collected$from$GPS$receivers$and$Stated$preference$survey$in$addi2on,$compares$cyclists’$preferred$ 2008$ Harvey$ Minneapolis$ routes$with$the$shortest?network$paths.$N:$49$ GPS$?$mounted$to$bicycle,$used$for$7$days.$Par2cipants$through$newspaper,$bike$shops$and$random$phone$ calls.$N:$164$ 2008$ Dill$ USA$?$Portland$ Switzerland$?$ GPS$?$data$collected$by$private$company$analysing$billboard$exposure.$N:2435$(people$in$original$survey).$ 2010$ Menghini$ Zurich$ N:$3387$stages$(taken$by$bike).$No$demographic$data$available.$$ Canada$?$ Montreal$ 2011$ Larsen$ Online$survey$?$travel$behaviour$analysis.$N:$2917$ USA$?$San$ Smartphone$GPS$receiver$and$applica2on$Cycle$Tracks.$Promoted$principally$through$bicycle$coali2ons$and$ 2011$ Hood$ Francisco$ university$networks,$but$also$media$outlets$and$local$events.$N:$952$ Mobile$applica2on$Rothaim$Dublin$(three$week$tes2ng$period$w$volunteers)$?$tracks$and$plans$routes$?$ programming$aspects$detailed.$N:$63$ 2012$ Gavin$ Ireland$?$Dublin$ 2012$ Hudson$ USA$?$Aus2n$ San$Francisco$CycleTracks$Applica2on$?$home/work$postcode?$(self$register).$N:$317$ Denmark$?$ Online$ques2onnaire/survey$asking$cyclists$to$map$posi2ve$and$nega2ve$experiences$on$their$most$recent$ Copenhagen$ bicycle$route$taken.$N:$398$ 2013$ Snizek$ 2013$ Van$Duppen$ NL$?$Utrecht$ GPS$traced$and$video$recorded$ride?alongs.$N:15$ Self?classifica2on$of$trip$taken,$the$data$collec2on$takes$accelerometer,$cell$and$gps$data$to$classify$routes.$ 2014$ Nitsche$ Austria$?$Vienna$ N:$15$trial$volunteers.$ Denmark$?$ GPS$volunteers$?$bicycle$route$choice$model.$Travel$diaries$from$select$par2cipants.$Data$collected$for$all$ Copenhagen$ modes$of$transport$and$post$processed.$Exact$details$for$data$collec2on$procedure$unclear.$N:139.$$ 2014$ Halldórsdólr$ 4 2015$ Yeboah$ UK$?$Newcastle$ GPS$volunteers$?$bicycle$route$choice$model.$Travel$diary$and$GPS$for$1$week.$N:$79$commuter$cyclists.$$
Data Capture Techniques (19) • Interview (1) • Ride-along interview (1) • Paper-based survey (5) • Online survey (2) • Purpose-configured GPS unit (5) • Smartphone (4) • Sensor configured bicycle (1) 5
Interview 6 N: 254 before, 110 after. (Lott, Tardiff and Lott, 1978)
Ride-along (+ GPS + Video) 7 N:15. (Van Duppen, Spierings, 2013)
Paper Travel Surveys • Travel diaries • Relies on good participant understanding of maps • Recall vs actual 8 Digitised hand drawn cycle routes in Guelph, n: 397 (Aultman-Hall 1997)
Online Travel Surveys • High response rate possible: n ~ 3000 (Larsen, 2011) • Respondents can contribute opinions/suggestions 9 398 online surveys – Copenhagen (Snizek et al. 2013)
Purpose Configured GPS • Often connected to PDA – allows user input (travel diary) • Bohte & Maat (2007) – self classifying journeys 10
(Kasemsuppakorn$&$Karimi$2013).$ 11
Smartphone application (Hudson$et$al.$2012)$ 12
• Modal classification based on accelerometer, GPS, cell network data • 95% accuracy for identifying cycling (Nitsche et al. 2013) 13
Sensor-configured bicycle (Eisenman$et$al.$2007)$ 14
Norwegian context • Development 8,0 by 2015 8,0 by 2023 on the ground • Big ambition! $$$ 2017 2020 2023 15
My research: • Office relocation. n: 114 • Bicycle: 10% (before) – 27% (expected after) • Bus: 8% (before) – 37% (expected after) • Parking fees and localisation (12km from CBD to 1km from CBD) Perceived$ Cycling$ Safety$ 16
Coming... 17
Where next? • People who do not cycle much – what would make them cycle more? • What about the non-commute trips • Network impacts of built environment changes (esp. Unique bicycle infrastructure) 18
Raymond.Pritchard@ntnu.no 19
Recommend
More recommend
