Could the Strava App Provide the Biking and Walking Data Cities Crave?
Strava may be making the leap from feel-good gadget for hard core exercise buffs to serious planning tool for cities looking to improve active transportation.
Strava is a mobile app that runners and cyclists use to record their activities, track their progress, and see their stats and personal records. Its website shows a bunch of young, fit, white people in performance gear, riding 12-ounce racing bikes in an aggressive position or running through woods with their feet never touching the ground. “Prove it,” seems to be their motto.
Rather than proving bragging rights about your latest personal record, Strava might be able to prove the need for safe cycling infrastructure on a heavily ridden route, or the increase in use since a protected bike lane was installed.
In addition to showing users the maps of their own routes, Strava is now releasing its global data online, creating heat maps of where people run and bike — or at least, where they run and bike with Strava tracking their movements. Users can separate out running and biking — the maps in this article only show biking.
Oregon DOT has caught on to the utility of this information and has become the first state transportation agency to pay Strava for its data. Jonathan Maus at Bike Portland reported yesterday that ODOT has paid $20,000 for a “one-year license of a dataset that includes the activities of about 17,700 riders and 400,000 individual bicycle trips totaling 5 million BMT (bicycle miles traveled) logged on Strava in 2013.”
This data matters, Maus says:
The problem for many transportation agencies today is that, while bicycling is on the rise (for both transportation and recreation), there remains a major lack of data. This gap in data makes it much harder to justify bicycle investments, plan for future bicycle traffic growth, illustrate the benefits of bike infrastructure investments, and so on. It also makes non-auto use of roads very easy for agencies to overlook. And while ODOT and many cities do bike counts already, they only measure one location for a short period of time.
Biking and walking advocates need good baseline data to prove the need for better infrastructure and to measure the benefits once that infrastructure is put in. That data can also help justify similar projects elsewhere. Will the Strava maps fill that need?
It’s hard to say. In all likelihood, Strava is primarily capturing data from recreational cyclists and runners, not people who commute on foot and by bike. Most Strava users probably choose routes for speed more than the convenient access to important destinations. But some utilitarian cyclists use trackers like Strava too, particularly for contests like the National Bike Challenge, which started yesterday. (Some note that they’ve ended up uninstalling it because of battery drain.) And though I don’t have demographic information on Strava users, my guess would be that they tilt wealthier and younger than the overall biking and running population of a city.
Still, the maps could help planners see heavily-cycled routes that are in need of infrastructure or note the streets that cyclists tend to avoid. The baseline data Strava provides would be littered with caveats and disclaimers, but it’s still more than we knew before, and presented in a compelling visual, zoomable to the finest-grain detail.
Other jurisdictions are trying to capture this information without Strava. In 2010, the San Francisco County Transportation Authority created the CycleTracks app that people could load onto their phones to let the city see where they’re going (with personally identifiable information stripped out, of course), what route they took, what the purpose of their trip was, and more. Planners say it’s helped them craft the city’s bike plan, with realistic data about how the city’s bicycle network was being used and what the needs were. Other cities have piggybacked off San Francisco’s success with the app and adopted it for their own use.
In addition to missing non-recreational users (commuters as well as uses for other purposes), I’d suggest it is also likely to be missing data from poorer users, underrepresenting lower-income communities.
Some great examples are the seeming lack of runners (and by extension: walkers) in Silver Spring and the Takoma / Langley / Long Branch area… these rank among the higher-volume ped/bike/transit areas in the suburban DC region, but look completely devoid of activity in the Strava map.
Good idea, but this particular implementation is questionable because it captures a small, self-selected set of users. What’s galling is that Oregon has chosen to get milked to the tune of $20,000 a year for this … sheesh.
I do not think this data is good for anything but planning your own route should you happen to be a fit recreational cyclist looking for a popular ride. Sometimes bad data can be worse than no data. For example, I suspect that near where I live on the UWS, there are probably more bikes on the Hudson River Greenway than on parallel Riverside Drive, but there are more Strava users on the latter. There is no shortcut to going out there and counting people on bikes.
The article said the map was bike-only.
A UK survery of 276 Strava users shows: The majority of cyclists identified as recreational cyclists (81%) and/or utility cyclists (70%). Only a few identified as competitive cyclists (14%). http://www.ucl.ac.uk/uclic/studying/taught-courses/distinction-projects/2012-theses/Williams-2012
To be clear, this was a survey, with a fairly low number of respondents. But the blanket assumption that Strava is only for male racer types is likely not accurate. Looking at the heatmap, there are clearly heavily traveled areas in cities, showing that it’s not all 20mph country road sufferfest rides.
There’s a “run” version too. http://labs.strava.com/heatmap
If someone uploads data and marks it as hike or walk, that wouldn’t show.
Strava is loaded with non-recreation riders. I logged every commute from the past year, as well as thousands of miles from recreational riding. Many of the people I follow on Strava post commutes.
Frequent cyclists know that riding on shared use paths is a drag. With walkers, dogs, strollers, rollerbladers, etc. it’s difficult to ride safely about 10mph on a shared path. That’s fine for a novice cyclist just looking to be outside, but most commuters will stay on the roads. Especially the strong and confident types will want to go 20mph on a road, not 10mph on a path.
20 grand for a huge set of data that included what times people ride during the day and which days of the week, and at what speed could be highly useful. A Strava employee commented about this elsewhere. With the data bought, they can look at where the fast riders go, and where the slow riders go. I’m sure a gov entity didn’t just write a check without bureaucratically arguing about its effectiveness for months until they decided that it really would be useful first.
Previously I used strava when I was training on my road bike. Once I started using my bike for everything else including commutes, Traders runs and meet ups with friends I haven’t used it at all. Most people I know who just ride their bikes don’t use it at all. After reading this article, now I’m thinking maybe I’ll start using it again.
Even if it captures commuters, it still has the problem of only collecting data on people with the money for Strava who choose that brand. In Richmond VA, the working-class southernmost third of the city and the areas near housing projects have nearly no cyclists on the Strava map. Using Strava for planning would be like planning car roadways based on GPS data only from GM cars built in the last three years.
Is Strava even the most popular app? Map my ride has way more reviews than Strava.
Strava is free to use, just need a GPS capable phone or other GPS device (another financial barrier but not limited to Strava specifically.)
Funny how everyone is dissing Strava’s App because they claim it only captures a certain subset of cyclists (which is admittedly, VERY likely) but Streetsblog and other planner types go gaga about cities like San Fran using similar tracking apps to do bike route planning and are then totally blind to the same reality that these tracking apps are only collecting the data of a certain tech savvy (and likely young and hip) cycling subset. I completely expect that these apps are totally missing the bicycle travel patterns of the poor and indigent and that their needs will be totally overlooked if planners start over-relying on smartphone cycling app data.
Never mind us Luddite types riding steel frames with dumbphones that then turn them off so we can’t be tracked. 😉
Civic hackers in Philadelphia, along with a regional planning authority, just released a smart phone app to track bike habits. It’s specifically asking for the types of ride, i.e. commute, recreation or whatnot.
http://technical.ly/philly/2014/04/30/cyclephilly-dvrpc-corey-acri/
It’s a project that’s also been run in Austin, Puget Sound, and Atlanta. And it’s all based on San Francisco’s code base.
The problems of the Strava data go way beyond only a certain type of person using them. For example, I went on the Five Boro Bike Tour in NYC yesterday, and the route for that tour is clear as day on the Strava map for NYC. There is a solid mass of Strava users going along the FDR Drive and the BQE to the Verrazanno Bridge. Taking any of those roads any other time of the year is akin to a death wish. Those are just a few clear-cut examples. Other roads may be more ambiguous, but the data is equally flawed.
That is categorically untrue. I could easily reach 20+ MPH during off-peak times on the greenway (with a strong tailwind). Strava users prefer Riverside because it is just a tad faster to the GWB, particularly for the return trip when more strollers and dog-walkers are out and one can run red lights without having to worry about cross traffic. It is clear this is their true motive because the Riverside line gets thicker as one gets closer to the bridge.
We had some discussion about this for my local Complete Streets group–the most heavily Strava’ed routes were some of the most car-heavy, while there were excellent side-streets (just one street over) that were generally agreed to be more bike friendly. While some of the information is useful, I’d take it with a large grain of salt.
The map for Boston shows a lot of Strava users on the bike paths and multi-use paths (and many lower-speed congested streets with heavy traffic, both bike and car). I was surprised by this – especially the high use of the SW corridor path, since that route is clogged with lower-speed commuters (and kids riding bikes) and there are numerous stop lights.
What’s more amazing to me – for Boston, is just how much of the suburban area of the city is covered by strava users – they’re everywhere. There does seem to be a slight drop-off in the lower-income areas of the city and suburbs, though… but at least for Boston, strava users do a pretty good job representing where people actually bike.
There is other better data. For example in DC Capital Bikeshare is probably a lot more indicative though you miss the long haul cyclists. This probably is useful outside of the core though.
I believe the bikeshare data just includes their start and end point. Maybe they make guesses on where the route goes, but I didn’t think any bikeshares actually tracked with GPS yet.
I think the point is that even though the SF data is still not representative, it at least represents people trying to get where they’re going and running their daily life, rather than occasional exercise and recreational activities.
$20,000 a year is really nothing for a government agency like this. That’s comparable to the cost of replacing one traffic light.
FYI Garmin has a similar heatmap for both running and bicycling that is similarly interesting. It appears to have less data at this time than the Strava heatmap so the caveats apply even more so. Nevertheless, very interesting and potentially useful.
Here is a sample:
http://connect.garmin.com/course/6488314
Click on the ‘layers’ button to select bicycle or running heatmap. Zoom in to see the specific roads or trails that are being used. Zoom out to see heatmaps from other places.
I randomly drew a route just for demo purposes that just happens to go near Garmin headquarters (near 151st and I-35 in Olathe, KS), so you can see some potential bias there as by far the highest use area is very, very near Garmin headquarters.
So are these roads good targets for bike lane improvements?
SFMTA’s app is certainly something newsworthy for Streetsblog, nothing “gaga” about that. Nothing “young and hip” about it, either.
find the purpose of life…just when I was about to give up…www.qly.us