2014-11-06 23:32:03
28
Strava data use
Does anyone if the city or BikePGH use ride data collected from services such as Strava or Map My Ride to determine rider tendencies? Obviously these services are geared more towards fitness enthusiasts than commuters, but encouraging riders to log their commutes on these services could be useful when mapping out future bike infrastructure.
Of course, there's an issue here with how companies use your personal data, such as what Strava is already doing (http://metro.strava.com). However, besides the fact that you agree to let them do this by signing up for their service, it appears that any ride/run that you post and mark as private would not be included in the data package (http://www.tuaw.com/2014/05/23/strava-begins-selling-your-data-points-in-the-hopes-of-creating/)
Thoughts?
2014-11-06 22:36:38
I think the City is considering the acquisition of this data for their use.
I saw it used on Westmoreland County in a planning application for the first time yesterday. It was cool, but confirmed the limitations that I had previously identified:
1. It identifies routes currently being used by STRAVA users, not by cyclists in general.
2. It tends to tell us what we already know -- that cyclists use a particular routing. Not why, or anything else.
3. It may show multiple routes being used between two points. The heat map/color density may show route preference (as indicated by higher use) but with only a limited population to pull from (Strava users) the variations in user density were pretty small, and not very helpful for planning purposes.
4. It DID demonstrate that cyclists were already using a route that was under consideration for improvement. That was helpful.
2014-11-07 10:27:42
@swalfoort, exactly. The system does have its limitations and would have to be used as a complement for data gathering, rather than being the main source.
One way to interpret this information is to show where cyclists *don't* go, which could in turn lead to the city investigating why that is (lack of infrastructure, hostile roads, etc.). And as you say on point 4, it can demonstrate that cyclists are using a particular route. This could be useful when evaluating the utilization of the new cycle tracks in the city.
Perhaps BikePGH could encourage commuters to log their rides (through any service) to help demonstrate the use of bike infrastructure in the city? I know that when you edit your rides on Strava you can tag the ride as a "Commute". This could make it easier to identify what kind of rides you're doing. BikePGH could even create a club on Strava, where you can see a leaderboard of who logs the most miles/bike trips. There could be some sort of recognition for being a top commuter
2014-11-07 11:21:02
Oregon DOT created their own app to collect bicycle data http://bikeportland.org/2014/11/07/odot-releases-orcycle-smartphone-app-collect-biking-data-113223
2014-11-08 09:20:47
A flaw with the Strava maps, that I pointed out in another thread ( http://localhost/mb/topic/crowdsourced-bike-map/#post-304109 ):
"A thought experiment: Suppose there are two trails, side by side, that receive equal numbers of cyclists per day. Trail 1 is straight and fast, so GPS devices there record one point every 500 feet. Trail 2 is sinuous and technical, so GPS devices there record one point every 50 feet. With Strava’s current heatmaps, Trail 2 will be displayed with 10x as many points as Trail 1, so it will appear to have 10x the volume of cyclists, even though the two trails are equally popular."
Strava has a service, Strava Metro, through which they'll sell their data points to you, so you could do your own analysis on it. With that, you could remove the bias toward slow cyclists that their current maps show, but two problems would remain: the data would still be biased toward the sort of person that uses Strava (i.e. competitive), and it would still contain (I imagine) some garbage data, e.g. user forgets to turn off recording on the Strava app or on their GPS device before getting in the car.
2014-11-08 18:08:39
I would like to know if there is a Strava map for Copenhagen.
2014-11-08 18:35:35
Copenhagen is here. But beware that these visualizations are time-weighted, so slow bikers are exaggerated relative to fast bikers. http://labs.strava.com/heatmap/#11/12.48523/55.68748/gray/bike
2014-11-08 20:11:39
Consider also the Denmark (Helsingør) to Sweden (Helsingborg) ferry connection, which seems to have bike traffic. And more interestingly the Öresund bridge/tunnel has bike traffic (or, people forget to turn their Strava thingies off).
2014-11-08 21:43:17
My understanding is that the City is intending to announce ahead of time that they're planning to collect the data, so as to push people towards using the system and be recorded. This would be a contrast to, say, Oregon's purchase of already-collected data (my own previous response: http://localhost/mb/topic/how-were-actually-using-the-roads/#post-295265), but I'm not sure if it's enough.
2014-11-10 18:30:48
Now it appears some other cities are developing their own apps to track this kind of data. Glasgow is the latest one:
http://www.cyclingweekly.co.uk/news/latest-news/mobile-glasgow-app-launched-144655
I'm guessing that there's a company out there that's developing custom apps for particular cities/states. Pittsburgh could do the same
2014-11-14 08:07:28
2014-11-14 10:13:23
I have been religiously logging every trip I've made by bike via Endomondo since the thing started in 2012. If I go back in and click on any given day, it shows a breakdown of what I entered, which I usually do from a pull-down menu I myself created. So I can tell, for example, whether I used Perry or Babcock or McKnight to get into town.
2014-11-15 10:41:42
I ran across this $80 DIY counter mentioned in a news article. It looks pretty robust and easy to build.
http://tomorrow-lab.com/lab16
2014-11-19 11:29:01
Strava has published the 2014 vs 2015 heatmap comparison. You simply slide back and forth the split line to see the difference between one year and the other:
http://labs.strava.com/heatmap/2014-2015.html#15/-79.99643/40.44435/gray/bike
There's definitely an increase in usage between 2014 and 2015, particularly on the routes that have gotten new bike infrastructure (Penn Ave, Bayard-Bigelow-O'Hara, 40th St, etc.). It also sees increase in areas that don't have bike infrastructure but are being considered for that, like BOTA. Of course, this may simply mean that more cyclists are using Strava, not an actual increase in bike traffic, but it serves to further quantify what the real numbers are (you have to pay to find that out, though)
2016-01-07 09:25:46
It was interesting to me that there was a notable increase on the far side of Mt Washington generally. Has the over the road environment improved there? The least stupid argument against bike infrastructure is that it serves too few people; hopefully that argument is starting to get undermined in more and more places.
2016-01-07 11:03:20
I think it's borderline irresponsible to make a map like this and not to provide a street name overlay.
2016-01-07 11:18:11
@jonawebb,if you drop down the map info section on the upper left hand corner, you can toggle on the street labels. You shouldn't have to, though.
Edit: just noticed that even if you toggle it on, it's not very detailed
2016-01-07 11:27:29
Worth noting: if you log your bike trips on Strava, but set them to "private" so only you can see them, they won't be used for generating their general heatmaps. So that 2-mile trip to work or school that you've been logging but keeping private won't be represented in this map. This algo goes for any privacy areas that you've set (a radius around a particular point, such as your home or place of work). Any trip data for those areas will not show up in the general heatmap.
2016-01-07 12:33:30
Cool.
One observation: Between 2014 and 2015 thin blue lines appears running along the rail tracks on the South Side and up Panther Hollow.
I assume they're for real, so who's riding the rails? (So to speak.)
2016-01-08 08:43:59
@Ahlir, that's most likely an error in the way that they set up the map, or an errant GPS unit. I noticed it, too, but I don't think it worth thinking much about.
That being said, I was on the Mon Wharf the other day, near the pillar for the old Wabash bridge, and saw a cyclist on (I think) an MTB heading east that did not come back. I figured said cyclist most likely went up the stairs on Wood and continued on to the Smithfield St bridge. But what if he continued on along the river? Hmmm.
Related to that, you can see that bike trips along the Mon Wharf have increased from 2014 to 2015. Other tidbits:
- An apparent significant increase in trips over the 16th St bridge
- The reemergence of the Railroad St + Strip District trail route in 2015. I expect this to be much denser in 2016
- The effect of reopening Penn Ave through Garfield and along the Allegheny Cemetery
- The use of the new sidewalk along Rt 28, between Penn Brewery and the 31st St Bridge
- An increase in popularity for Brookline Blvd (perhaps due in part to West Liberty Cycles moving there?)
- Riding through Uptown and the Hill District remains virtually unchanged
2016-01-08 10:19:57
I don't think you can get upriver from the Mon Wharf. I tried it once, and gave up after about 100 yards. In the past, I have managed to get upriver from Duck Hollow to Carrie Furnace staying very close to water level, and believe me, that wasn't easy. But from the Mon Wharf it was well nigh impossible.
2016-01-08 10:51:39
There was little change in uptown, but there was an uptick on Forbes/Fifth in Oakland. BRT and associated lanes will be a game changer if done credibly whenever it does come. This would provide a nice visualization to see that. Similarly if the city could get its act together on the Wabash tunnel or something. Any updates on either?
The gap in the Hill district is saddening to me. It could get quite a bit better with decent facilities on Center. Of course the big enchilada would be a funicular or bike lift coming up from the strip, though I'm not too sure there'd be much overlap between that and the population of Strava users.
2016-01-08 10:55:10
Center (or Wylie or Webster) is not actually that bad to ride on, if you don't mind hills -- at least, when I've been on it, on weekends. It's wide, so traffic isn't too much of a problem.
The big problem with riding in the Hill District is, well, the hills.
2016-01-08 11:01:20
It's a somewhat different thing, for folks who are less road warrior-ish, to be riding where there's a bike lane vs. a wide lane that's often, but not always share-able. Even if traffic is low (which I've also found true on Center), the psychology of it is a major hurdle. Especially if a pinch point occurs going uphill. Though I do agree in the large, there's also a limit how much better things get without an easier way (physically) to clear the necessary vertical coming back to the neighborhood.
2016-01-08 11:30:58
Precisely. There's little incentive to use the Hill District as a through route. It involves a hill climb each way, regardless of direction of travel. So even if Centre/Wylie/Webster Aves are somewhat pleasant to ride on (at least when compared to Forbes and Fifth), it remains an unappealing route in general.
As for the people that live there, if you work downtown, the Strip or on the South Side, riding to work makes total sense to me. Downhill all the way! To Oakland/East End, you'll have to climb a bit before heading down, but it still isn't too bad. Coming back, Centre is not a bad climb, and Kirkpatrick is relatively wide and pleasant to ride on, at least from a traffic perspective. Brereton St + Herron Ave, or Herron Ave all the way, is a tough climb any way you slice it.
I can't remember who told me this, but apparently in the past there was little or no interest from Hill District leaders/residents for installing bike infrastructure there, but that has changed recently. It would be interesting to see if the City has any bike infrastructure plans for this area (that don't involve the Civic Arena site).
2016-01-08 11:32:08
that’s most likely an error in the way that they set up the map, or an errant GPS unit.
It's not errors: the traces are right on top of the tracks (on the map) and they relate to the adjoining bike/ped trails just like in real life. GPS error is seen as the blur on streets. And you need a lot of repeat traffic for stuff to show up on Strava maps.
And now that I had to look again: Penn Ave, with the bike lane is suddenly very much more traveled! Spring Way is more popular, something that's been noticeable on the ground and I think goes with the advent of the Penn Ave lanes.
2016-01-08 12:51:21
@Ahlir, the same can be said about all of the infrastructure installed in 2014 and earlier in 2015 (Schenley Dr, Saline St, 40th St, 6th St and Clemente Bridge, etc). The increase in trips on Spring Way because of the Penn Ave cycle track definitely applies to me, too.
Another one: increase in traffic through the gas station in Station Square, likely because of the shortcut closure near the railroad tracks for the South Side trail. Expect this line to be denser in 2016
2016-01-08 15:20:05