PSRC wants to know where and why Puget Sound bicyclists ride, and they have partnered with Bay Area smartphone app Cycle Tracks to allow bicyclists to record their trips and share their data directly with PSRC. The data will help prioritize present and future investments in bicycle infrastructure and will improve the ability of planners to model cycling demand variables such as slope, arterial speed, presence or absence of bike lanes, etc…

Cyclists may start using the app anytime, while PSRC will begin collecting data on May 1.  The app is available only for iPhone (download) and Android (download).  The app asks for your age, email, gender, home/work/school ZIP, and cycling frequency, but all of these inputs are voluntary.  For the purpose of grouping trips by user, the app does require your Unique Device Identifier, but this data will not be shared.

I have been testing out the app for the past week, and it has a generally pleasant user interface and its basic functions work well.  Trips are saved by date and trip purpose, all trips are mapped, and average speed is given. Other apps such as SpeedTracker provide more information, such as color-coding your map by speed so you can see your bottlenecks. There are some unfortunate quirks, such as the app only being able to run on an active (though dimmed) screen (i.e. you can’t press the top button on an iPhone or this will stop the trip from being recorded)  Using CycleTracks on my 13-mile bike commute drains about 35% of the battery.

Overall, this app represents a nice chance for cyclists to allow their own travel preferences to directly impact how planners understand cycling.  An app FAQ is here, and questions can be referred to Peter Schmiedeskamp,

36 Replies to “PSRC Wants Your Bike Data”

  1. The app is available for iPhone and Android only.

    *shakes fist*

    1. Sorry! I assume you’re looking for a Windows Phone or Blackberry version? Unfortunately, we don’t really have the resources to port the application to other platforms.

      In particular, I would have liked to see a Windows Phone version if for no other reason than to capture route information from the scores of hard-bicycling folks at the Microsoft campus!

  2. Total fail on the post, not a single direct app store or google market link in the article or linked articles. yet you included a link to a competing app.

    If your goal was to get people not to use it, you did a great job.

    1. I thought the link in the first paragraph was sufficient, but I apologize for not being clearer. I added download links in paragraph 2.

    2. Anyone too lazy to type two words into a search bar is probably too lazy to hit the ‘on’ button when they ride anyway…

  3. If the goal is to collect more data, it would also be useful to include ways to upload the data I’m already tracking (e.g. from my Garmin bike GPS). I’m not likely to run an app on my (android) phone in addition to my GPS (especially if needs to keep the screen on).

    1. That gives me an idea. I was planning on switching from my usual app (MapMyRide) to this, and was sad that my data would be in two places. I wonder if I could run both apps at once…

      I actually use a handlebar mount with my iPhone, which I recommend. Not only do I have a GPS in front of me for unfamiliar trips, but if I get a call I can look to see if it’s someone worth stopping for.

    2. Hi Scott,

      From our (PSRC’s) perspective, the more data the better. Unfortunately, we’re not really set up to collect the other bits of information such as trip purpose and demographics in a consistent way from the Garmin devices. Part of the beauty of the CycleTracks application for us, is that you record your trip purpose while the ride is fresh in your mind.

      Sounds like this will be a good topic for me to add to our FAQ!

      1. Part of the beauty of the CycleTracks application for us, is that you record your trip purpose while the ride is fresh in your mind.

        This makes sense. A week later I often have no idea why I biked to work at 9 AM, why I biked home at 6 PM, why I biked to the grocery store at 7 PM, why I biked to a friend’s house at 8 PM, or why I biked home at 10 PM. That’s information I definitely couldn’t infer given the route data.


  4. This is also the time of year that mass emails start going out for Bike to Work month. I’d suggest passing this along to any team captains or communications people at your job.

    1. We’d be much obliged if folks did! Personally, I’m planning to use CycleTracks to keep track of my trips for Bike to Work month. It should keep me honest on my mileage if nothing else.

  5. Holy biased data, Batman!

    People’s cycling patterns are largely determined by cycling conditions anyway. Create good cycling conditions in places people want to go and some people will bike there. Only a few will have phone mounts on their bikes, either way.

      1. Bad data == bad conclusions.

        Besides who really wants to put their $600 iPhone out in the rain to collect data for this?

      2. Wait a second, why do they want data on where we currently ride? I currently pick streets that are “safe” not necessarily because they are direct. Wouldn’t it be better to collect start, and end data and any other destination to destination data? Seriously they don’t plan freeways this way… gosh Homer drives down this dirt road to the next town, then over to that farm and then up to the town over the hill… No they say, how can we get the most people who start at A to B.

        This study is seriously stupid.

      3. 1. You don’t need a bike mount or to put your phone in the rain. Turn it on, and put it in your pocket. Or pannier. Or wherever your phone is supposed to go. I prefer a water resistant bike mount, but it’s far from necessary.

        2. Knowing how you use your bike is useful. If we find that we spend a lot of effort on sharrows and bike lanes on arterials, but everyone rides in the residential streets because they’re safer, that’s valuable information that can help agencies change strategies.

      4. Gary is right, Matt and Chris are wrong.

        If you collect “some data”, all from people inclined to download an app and play with their phones surrounding their commutes, you will receive a biased picture of where people ride. You can’t correct for this just by collecting demographic information, as filtering by people that have smartphones, have heard of this study, and are willing to participate captures a distinct demographic in and of itself.

        Seattle is marginally bikable. This isn’t for lack of knowledge or engineering ability. It’s a political problem. It’s for lack of political will to slow down and restrict car traffic. It’s for the unwillingness to treat cycling as a form of transportation rather than only a form of recreation in political discourse and planning documents. It’s this final barrier that I think we’re finally breaking down to some degree; that’s the path to real progress.

      5. Planning is not the same as science. If you were in the pitch dark, you could learn a lot more about your surroundings with a flash light that only pointed one direction than you could with none at all.

      6. We’re not in the pitch dark. We have transportation experience from many transportation projects, for better or worse. We have examples of good bike infrastructure, mostly in other cities and other countries. We merely lack the political will to make the necessary changes.

        The Dutch didn’t even have examples like this when they turned their cities into the best biking cities in the world. They did it because they wanted to, not because they knew how. Biking isn’t rocket science. Even the bits that aren’t common-sense are easy enough to figure out.

      7. why do they want data on where we currently ride? I currently pick streets that are “safe” not necessarily because they are direct. Wouldn’t it be better to collect start, and end data

        Gosh Homer, if they collect “trip” information, guess what, they have your start and end data. DOH! By looking at the route data graphically it will be obvious where the popular end nodes are and where there are large detours from the most direct route. We already know where employment centers and high density exist so a lack of trips to/from one of these relative to another speaks volumes about the need for a “missing link”. It’s not nuclear power plant science ;-)

    1. Zowie! First things first, you don’t need a phone mount. On my iPhone, I simply start recording my trip, and then throw it in my pocket or pannier. Looking at the data we get back, I can confirm this works just fine.

      As for bias, we’re definitely well aware that this data comes from a self-selected sample of smartphone owners. It’s an issue, but not as big an issue as you might suspect. Some things to note:

      1. We’re NOT using these data for bicycle counts! The purpose of this study isn’t to figure out which routes are popular, it’s to figure out how far out of their way bicyclists go.

      If we didn’t know *anything* about bicyclist travel, we might assume that bicyclists would simply take the shortest path. However, we DO know that bicyclists choose longer routes based on lots of criteria (e.g. hills, traffic, bike lanes, etc.). By measuring the departure from the shortest path, and marrying it up with demographic, topological, and other route attributes information, we can begin to quantify what makes for “good cycling conditions”–and with luck, know how to build more of them.

      If you don’t mind geeking out a bit, I recommend reading San Francisco Transportation Authority’s CycleTracks study. Part of what I like about our study, is that we’re not reinventing the wheel. If you want to know roughly what our study will look like, this is probably the best place to go. Kudos to SFCTA for creating a reusable data collection platform for agencies like ours to use!

      2. In terms of actual route choice, we don’t really think that smartphone ownership is much of an influence. For example, bicyclists don’t avoid hills because they have smart phones do they?

      3. Other factors that we DO think will influence route choice such as age, gender, and cycling frequency are explicitly included as variables.

      In short, sample bias is a problem, but not a fatal one.

      1. Cyclists may not avoid hills because they have smart phones (maybe — I actually wouldn’t rule out hill behavior changes correlated to smart phone ownership). But this isn’t just people with smart phones. It’s people with smart phones that are plugged in enough to cycling/transportation advocacy to hear of your study, are technophilic enough that they want to fiddle with their phones when they get on and off their bikes (FWIW I write software for a living; the last thing I want is another digital distraction). Do you think that group might be different from the general population in how it reacts to hills? To traffic? To crime? I think there’s a question of attitudes of participants that goes way beyond smartphone ownership.

        Furthermore you’re not sampling the people that aren’t biking. If we were to achieve Dutch levels of cycling, we’d do it more on the backs of current non-cyclists than current cyclists.

        And you’re not looking at facilities and techniques that our region doesn’t have, facilities and techniques that clearly have an effect on cycling elsewhere in the world. For example, we have no actual bike paths in this region, only multi-use paths, some of which have draconian speed limits and most of which are frequented by people that don’t feel the need to follow basic rules of traffic (this is our grim compromise; we can’t build good bike routes or good walking routes because we try to combine them into one facility even where this makes no sense). We have essentially no bike facilities with rationally designed intersections and yielding patterns. European urban cyclists may not be racers, but they have facilities for bikes alone, where people basically ride in an orderly fashion, and where traffic flows in a pretty predictable way. We have no bike facilities across which vehicle turning restrictions are limited — this is a nearly universal feature of bike facilities in Europe; don’t you think that having drivers stick their noses out in your right-of-way might discourage biking? Our best and most direct transportation ROWs with the best grades were grabbed by freeways, so in many places we don’t have the option of riding in the corridor with the flattest grade in a straightforward way. And if a freewayside path was built, freewayside development patterns are hardly bike-friendly, and freeway noise and air quality is hardly pleasant.

        The only thing you can measure is whether people prefer a door-zone bike lane on an arterial road (a bike lane that disappears the second you enter a congested part of the city, with useless signage if any), a multi-use path with confusing, no-visibility intersections and frustrating traffic conditions that goes where a train needed to go in the 19th century instead of where people need to go in the 21st, or a side street with poor arterial crossings. Considering what the world outside the US has to offer, this is a grave bias.

        The Dutch built the best biking cities in the world because they wanted to, not because they knew how.

      2. (The SF report is evidence that this whole enterprise is useless. A whole big study on urban cycling that sees only a choice between bike lanes and some kind of shared lane, and advocates for bike lanes. That contains lots of statistical jargon but not one instance of the word “door”. Not a single mention of intersection treatments. It can’t even study bike boulevards because they don’t exist yet. Maybe if we build something great in the physical world a study like this might be useful in telling us how well we did.)

      3. The problem is that your data collection is still waaaaay too limited. You should have an alternative means of sending the data you want rather than a GPS based phone.

        For instance I don’t own a smart phone (gasp!) I’d rather spend my money on the bicycle etc. Yet I’d being a transportation fanatic, I’d be glad to tell you where I ride. There are a number of ways for me to map that route ( etc.) And since I commute, it would be easy to say which days I rode which routes.

        But as Al has pointed out, who cares what route I take today? I ride these routes because they are the safe way to get to work, and home not because they are fast (it’s not, it takes me an extra 10 minutes to do this), it’s flat (its not, but safety over hills any day.)

        And why not work with a bicycling group like Cascade? Have them collect this data via their bike to work month program? They all ready collect distance and days, all you need is to attach that to a map of a route to know who’s riding where.

        But like Al says this isn’t really a data collection problem it’s a political problem. The road killing car associates are terrified that there might be a cheaper and healthier alternative to driving and spend an inordinate amount of political power and media effort to portray bicyclists as tax avoiders getting a free ride, and that they slow everyone else down. Just look at the comments on the Seattle Times boards whenever they write about a road diet. OMG you’d think that the world had ended and all commerce would end.

        As for locating a bicycle Blvd, that’s pretty easy to start, find a school, find an adjacent neighborhoods, and now make the streets to that school, 20mph, not direct through streets (ie force drivers to turn right, then left to get through) Berkley has done this, Eugene OR has done it, Portland is doing it. It’s a big no duh.

    1. (that’s not a public link – you have to be logged in and linked to John’s account as a friend)

  6. Well, OK, my phone’s not smart, but if they are trying to figure out where we need a better bike route, I vote for along Alaskan Way/Highway 99 between the West Seattle Bridge and 1st Avenue Bridge.

  7. What they should do is accept formats from other tracking programs. I’ve been using Google’s My Tracks ever since I got my Android, and it exports to different formats (doesn’t drain the battery and allows the screen to be off). I’d gladly upload the outputs from that if they’d accommodate it.

    1. I wish we could, but we’re also trying to ensure consistency in the survey instrument. Whereas we could probably match up your route information, we’d be missing your demographic information and trip purpose.

      I hope you’ll consider using CycleTracks–even if only for the month of May! The more data we can collect, the better the picture we can get of bicycling in our region.

      1. I totally fail to understand why you need demographic information AT ALL, unless you are selling this information to marketing firms.

        What difference does it make over what route a person takes based on their “demographic information?” Seriously, in an age where way too much information is collected about us without a warrant. Does it matter whether a driver on the Freeway is 16 or 40 or 70? No their car takes up just as much space as anyone else driving.

        What you need is ROUTE information. Where do people start, where do they go? And you need to know WHY they can ride, or WHY they can’t. (or perceive that they can’t.) Time information might be nice, but at today’s level of riding (3.6 % of commuters) congestion on bike routes isn’t an issue. Which means you need to know the routes of bicycle riders who don’t ride. That data won’t be collected by your app.

  8. Gary,

    They want demographic data because there is a demonstrated demographic difference in who cycles and, far more importantly, WHERE they cycle. Having relevant local information will only help when promoting new projects to the local politicos and public.

    As for selling demographic information, without the actual user’s name and contact info, it is kind of useless.

Comments are closed.