« Adult bicycle education EOTR | Main | Wednesday Morning Commute - K Street Rules »


Feed You can follow this conversation by subscribing to the comment feed for this post.

Nobody goes there anymore. It's too crowded.

Oops. Should have read the whole post BEFORE I opened my mouth. Makes me look like an idiot twice.

There is a simple solution for CaBi commuters: now that you recognize the awesomeness of getting to work by bike, get your employer to provide secure bike parking and ride your own bike to work. It doesn't devalue your CaBi membership since CaBi is still the best option to go places where there is no guarantee of secure bike parking. And those days your station is stocked when it needs to be you can leave the bike you own at home.

I think most of the congestion with CaBi is in the NW part of DC. I use CaBi to finish my commute by VRE and have had no problems yet. Luckily there is a good deal of flexibility in the system. For example in the morning if the rack at L'Enfant is full I stay on the train until Union Station. In the afternoon I check the status on CaBi's website and ride to which ever station has more open racks. I think expansion of the system this summer to fill in NW DC will help to clear up some of the problems.

CaBi causes too much congestion. Here I am trying to maintain a decent speed of 20 mph on the sidewalk of the Roosevelt Bridge getting to work. I get stuck behind this guy on a 50 pound red monstrosity. Its two little red lights were slowly blinking as if to say tauntingly "there's no place to pass, there's no place to pass..." I can't deal with this before my morning coffee.

Rush Hour is really the only issue, and only at a limited number of CaBi stations (though, obviously, the most popular ones). I've slightly adjusted my commute times and have only occaisionally had issues with an empty station when I needed a bike or a full station when I needed to return one.

We are definitely in the "growing pains" period of the system and, seeing how CaBi has responded so far (moving docks, hiring more rebalancing vans, etc.) I fairly confident this period will be relatively short ... especially if Wells proposal goes through (which I am not that confident about).

I posted about this on the WABA forum so I suppose I'm simply piling on here but I think that there are two separate issues here. First, as has been noted quite rightly by others it's bike SHARE not bike rental. If people want the assurance of a bike being available for their daily commute at 8 am sharp they should get their own bike out and ride to work. The whole premise of the system is that you take your chances on the stocking of any given dock at any given moment. I'm sure that as data is accumulated that improvements can be made on restocking docks if the effort gets made. There in lies the other rub. I'm not entirely sure that the effort is being made. I have more than a few times seen several docks in mixed residential / business areas in NW ( Dupont Circle, Admo, Mt Pleasant for the most part ) completely full with bikes at 11 pm on a week night only to discover that same dock still full at 6 am. I am hard pressed to think that CaBi had no opportunity to rebalance those bikes between 11 pm and 6 am when both bike and vehicle traffic was low. Is this an issue at the bottom of the resource pyramid of not enough vans and bike movers or at the top with insufficient time and attention from the people tasked with directing the rebalancing?

@Mase- rush hour is not the only time that the system doesn't work properly. Weekends are tough, and I saw a group of 5 cyclists outside soviet safeway at 7:30 last night hoping some slots would open.

@Riley- you hit on an important distinction- it's share not rental. That's the line staff at CaBi continues to use, stating that no one is guaranteed a bike or an empty station to return it to. Unfortunately, DDOT and Arlington County sing a different tune, citing it as an additional mode of transit. It would be great if they would all get on the same page in order to better manage expectations of users. I think people would be far less annoyed if they didn't feel like they were sold on something different than what we've got.

With that said, I'm curious how many time CaBi has been fined for full/empty stations that have not been serviced within the alloted time agreed to in the contract?

Not to be snarky, but I think my 4 year old might need to explain the concept of "sharing" to some of the CaBi detractors.

Ok, that was a bit snarky.

CaBi needs to fix this problem with the popular stations before it adds new stations. e.g. add more racks/bikes.

People will tolerate growing pains, but if you degrade the brand then its hard to repair.

CaBi causes too much congestion. Here I am trying to maintain a decent speed of 20 mph on the sidewalk of the Roosevelt Bridge getting to work. I get stuck behind this guy on a 50 pound red monstrosity. Its two little red lights were slowly blinking as if to say tauntingly "there's no place to pass, there's no place to pass..." I can't deal with this before my morning coffee.

You should use the time to do something productive, like work on your wheelies.

Hopefully the new stations this year (and in the future?) will help to smooth out the bike/dock availability issues.

Cabi is like a gateway drug. It gets people to try biking. When they see how easy it is, they will move on to their own bikes.

My great regret in life, Oboe, is that I've been unable to master the "wheelie" since the passing of my beloved Stingray.

Wow. I get mentioned twice on Washcycle.

OK, that probably wasn't me this morning blocking traffic. I didn't even know there was anyone behind me.

And the problem at the Soviet Safeway was Spotcycle was reporting the station as having several spots open, but in reality they were not available because of open bikes.

I didn't mind,but the GF was NOT happy. Wrote a nasty letter to bikeshare today. They are doing something funny with bike replenishing there, because I never had a problem at that station (I usually get there around 7ish) until a month ago.

Anyway, to focus on profitability for a second. Bikeshare in DC might very well become "profitable" because of day memberships and advertising. Not sure if that will apply to Arlington (less tourists and smaller ad money split).

But Alta may be losing money. I think someone on here said Alta expects costs of 155 a month, and the contract is for 150.

In addition, you've got to remember the capital costs of the stations.

I already bike commute but want to use CaBi when I want to/ have to / or feel like it. That's why it is there, isn't it?

Our building's gym is closed for renovation and we get to use a gym a couple of blocks away. I could easily walk there but it's a hassle on bike shoes.

Ideal solution: Make use of the CaBi stations just outside my office and the building I need to go to.

Today is the 8th day of doing this and since Friday, I have had four consecutive days where I could not find an open spot to drop the bike. Yesterday, I took a bike out of the full station on the return trip and was lucky to find the last open spot at the destination.

That is just bad logistics. Disappointing.

At that rate, it is not clear why I need this membership.

Eric's comment is exactly the problem. Here you have someone who should definitely take to CaBi, but is being turned off because of the mismatch between bikes and racks. We need to get more racks in the existing spots, rather than spreading it out more.

For an analogous situation, look at other forms of public transit. The data I have seen is that people are more motiviated to take the bus if it comes frequently and reliably, and so it is better to have one well-served route than lots of infrequent routes on parallel streets.

Another way to address the crowding problem is variable pricing. For example, if you have a dock of 10, make the last two twice as expensive, or only available to certain users who pay a premium for access. A two tiered system would distinguish between the regular users who need the service to replace regular commuting options.

Reading the PoP blog, you can see some seriously PO'd people. e.g. had to ride past three CaBi stations to find one open, which defeated the purpose of CaBi.

The problem is not that existing stations do not have enough docks, but rather that stations are so far apart. If you want to originate a ride from the station closet to you, you will give up when you find that the station is empty, because the nearest station may very well be a 10+ minute walk away. Similarly, if you want to terminate a ride at the station closest to your destination, you may be in for an unpleasant surprise when you discover that your ride is but half over. Oh, and after you finish that ride you will be doing a lot of walking to get to your actual destination.

If you look at the other large systems out there (London, Montreal and Barcelona), the stations are packed very tightly together. If one station is full or empty, there is almost always another station a block or two away. In DC, there are some stations that are more than half a mile away from closest other station. Virtually all of the stations are more than 2 blocks from any other station. With this design, the system cannot fail gracefully.

The fundamental error CaBi has made is to build out so far with so few stations in place. From the looks of it, CaBi covers roughly the same area as London's system with 1/4 as many stations and bikes. In practice, this encourages planned use (commuting) over spontaneous quick trips that everyone seems to believe is what CaBi was designed for.

KLO is right, the problem is that the system is not dense enough. There should be one every block, not 10 minutes away.

The area CABI covers in DC proper should have at least 200 stations.

EOTR, Crystal City and Arlington should be considered independent systems, from a planning standpoint. Same with that university up in the northeast (which has 3)

You know why nobody rides the EOTR stations? No, it's not because residents there hate bikes, its' because the distribution sucks. The area should have at least 25 stations, so that its actually useful (home->supermarket, metro->restaurant, etc) Right now it's "walk 15 minutes to get a bike to take to downtown" instead of "use a bike to run errands in your community"

Mind you, I understand that theres a huge pull to expand the coverage area. Not only did CABI want to hit as many wards as possible, but they want to reach a wider audience by saying "there are bikes in your neighborhood".

Unfortunately, thats not the best idea.

I'd like to see their Boston map. 60 stations for an initial launch is not good at all (the RFP called for 150). Im worried that they will try to spread out the stations, so that there are no calls of racism and classism, and in the process, screw everyone over.

Mr. Share, it would be better if the system covered the same area with more stations. But that wasn't an option. They had 100 stations to play with and for now I think they did a pretty good job. 100 stations at the density you're talking about would only cover the CBD. I can't see a system like that get 5000 users per day.

What do you think the boundaries of a 100 station system should be in DC?

Yeah, I think that it's important to acknowledge the political realities of station distribution. And I don't mean smokey backroom deals, but the importance of distribution of a limited amount of stations throughout DC. While 2-block topography might be ideal from a balance/service perspective, it's going to result in a lot of bitter and unserved population. And while that's sometimes just the hit you have to take, sometimes it's just not worth it.

As much as I like and understand the reasons for the dense distribution model of CaBi, I support a couple of stations along Columbia Pike as a bridge of sorts between the (assured) R-B corridor stations and the (existing) Crystal City stations.

Would it be more practical, from an operational standpoint, if we just filled in Arlington cell by cell? Sure. But that means it's years of Columbia Pike looking on while R-B gets all the benefits (as usual). That's just not helpful in the bigger picture.

I don't want to be overly negative. Capital Bikeshare has been successful, and I happily use it. I just want to point out that many of its problems are caused by the severe lack of station density. This problem can be ameliorated by placing all new stations close to existing stations, and I hope that is what DDOT does.

I also wanted to point out that a spread out system is not reliable enough for spontaneous trips. Many people seem to think that low reliability will discourage commuters, which will, in turn, encourage more random, spur-of-the moment style trips. Nothing could be further from the truth. The less reliable the system is the more planning one needs to do in order to use it. Commuters who travel the same routes every day are in a much better position to use an unreliable bikesharing system than are people who need to meet somebody for lunch one Saturday.

I have over 100 trips so far, and only 4 of them involved a full or empty station. That's pretty reliable.

MB, the problem is there ARE smokey backroom deals. It's standard DC bullshi*t politics. Every DC council member demands docks in their ward. KLO & Mr Share have hit the nail on the head. The density here sucks compared to every other bike sharing system worldwide. DC places stations everywhere from upper NW, to Brookland, to Anacostia to Georgetown, then drives diesel trucks everywhere to move bikes around? Not smart - we could have a more compact system located where a majority of DC residents live and work. Planners chose otherwise, and we have what we have.

Yes, our density is lower, but our coverage area is higher. I know I sound like a broken record, but we really don't know what the ideal station density is for any system. And it would differ based on your goals and the layout of your city. Maximum reliability is one goal. Maximum coverage is another. The two, in this case are in conflict. But I think coverage gets you more trips than reliability does.

I feel compelled to note that commuting from 16th st heights to the hill every day has had no trips where I couldn't get a bike and exactly one trip into work where I had to continue from my preferred station to the next one.

I know not getting a bike is frustrating but the system has become quite popular and obviously works well for many, many people based on the number of riders I see around.

Similar to Chris Adams, I have commuted from Georgetown to Judiciary Square about once a week (and home 2-3 times a week) for the last several months, and in 70 rides, I've had availability problems about 5 or 6 times.

I would argue against the density-at-all-costs approach, noting that the people that have the most problems with availability are also the ones with the most stations available to them density-wise.

More station density brings more potential users, which leads to capacity problems. There is a point at which you can saturate the market with stations (and not have the overload) but we're a ways away from that point. In the meantime, adding capacity in the core is good, but so is adding capacity in areas that are currently served, but in a less-dense fashion (like the Wisconsin, Connecticut, and Georgia Ave corridors).

Adding stations in those areas, and particularly adding stations about 4-5 blocks away from metro stations can turn existing "commute from" stations into "mixed use" stations. You see some of this happening at Dupont Metro and Eastern Market Metro stations now, which are just as likely to be used as ways to get from home to Metro as they are to get from Eastern Market to Downtown as an alternative to Metro.

In adding 25 stations this summer (and if the Wells plan goes through, 40 more in the following year), there are a lot of things to balance, but "thickening the trunk" at the expense of everything else has some serious drawbacks.

In response to washcycle's anecdote: I have over 300 trips since joining CaBi. I'd say that during the first 200, it was very rare for there to be a full or empty station. The next 100, it was more common to run into the empty or full rack problem than not (less than 50% reliability). I recognize that I find a lot of utility in the system despite this given how much I still seek to use it. But I also agree with those who think aggressive expansion in the core is very important before many users view the system as unreliable.

Many of the experts on biking, urban and transportation planning, logistics, the Myth of Sisyphus, etc., make some excellent points. However, remember prior to last September when Capital Bike Share wasn't even an option? Remember the tiny little bike share bikes that maybe 12 people used. Remember taking the bus or Metro trains, which, by the way, are still options? What did everybody do back in the day? Oh yeah, we all managed well enough without one extra outstanding, if imperfect option.

The fundamental error CaBi has made is to build out so far with so few stations in place.

Just want to chime in here that, with the system being not yet one year old, it's not clear CaBi management *did* make a fundamental error. They've got tons of members, you see people everywhere, and there's a Hell of a lot of demand that's not currently being satisfied--which means they've got the potential for political leverage to expand the system.

KLO is right, Bikeshare doesnt work well for daily commuting. Its about taking bikes for random trips. When commuting is the number 1 use, the system fails unless a new distribution model is deployed. For example, instead of having the truck be deployed after a station is full for x amount of time, have employees stand by the station and check bikes in directly to the truck.

Washcycle, I would have started by not putting a single station west of the zoo, excluding georgetown. The placement of those 8 stations is ridiculous. The only exception would have been stations for zoo visitors....and they have none! East of the zoo, nothing north of columbia heights metro. 14th and Decatur? What the hell is that? Could you imagine getting there and the rack is full? Youre 100% screwed.

The CUA stations are also odd, Ive never been to their campus, but they dont seem to be located in a way that allows good student circulation.

So thats 16 stations I would not have used for the core. Areas like Logan Circle and Thomas Circle which have NOTHING.

Dupont circle shouldnt have one big station, it should have three on different "corners"

EOTR, I understand it's a political necessity to include them. I would have used the same amount of stations but clustered them so they are actually useful. Right now, theyre just located in places that look good on a map.

Crystal City is excellent.

Arlington, I would not have started with less than 10 stations.

Heres an example of the type of density they should have gone for, ecobici in mexico city.

My computer isnt loading their google map image, so heres the PDF


You can stand at one station and see 3 others.

The http://oobrien.com map is a great way to see how the system densities compare.


Rio is a good example of "what not to do"

So you think there are 16 stations (I counted 17, 8 west of the zoo, 3 at CUA and 6 north of Columbia Heights) that are too far out. Pulling them in would increase density a little bit, but you're really only adding 1 station per square mile in your newly defined area. That's hardly an indictment of the density they went with.

Among the ones you singled out are ones that are pretty highly used. #14, #22, #43, #46, and #55 in total trips. And it includes the station PoP is complaining about.

A year in, Ecobici is getting about 9 tpbpd. We're at 5 after 6 months. And Mexico City is a bit denser than DC as I understand it, so they have a big advantage. It looks like they put them all in a strip about 3 miles long and one mile wide.

I'd love to know how many car trips their system replaces, compared to walking. I'd take fewer tpbpd if it removed more car trips.

Less stations with more bikes/racks per station would be cheaper to operate and, in theory, more reliable. With time, they will work out where they need big stations, and where they need smaller ones. Right now, I can tell that pretty much all the stations near govt offices are empty.

Washcycle, a few important notes about ecobici:
-Its a clearchannel system (ew) which means NO day users. Just annual members. So no tourists, no curious trials, nothing. You're in or your not. That makes their ridership even more impressive.
-Yes, Mexico is denser and more populated....but it's mexico city. I for one would not be comfortable riding my bike there. They did however install some separated bike lanes with deployment.
-It's relatively expensive. Annual fee is 300pesos, which is like $25. Much less than DC....but DC metro STARTS at $1.50 a ride. Mexico? Unlimited subway rides for 22 american cents. So compared to other transport, it's pricey. Cabs are also ridiculously cheap. I took a 20 minute cab ride (unsure of distance) for $3.

So all those strikes against the system (the biggest being it's clear channel) and they're doing better. I think the density is a huge part of it.

No idea on how many car trips it replaced. That's really hard to calculate.

Mr. Share, there are way too many variables here.

1. It's been open twice as long
2. Mexico City is denser - obviously you can build a denser system in a denser city. I think you've breezed over this way too much. This is THE critical point.
3. Mexico City has mild weather all year long.
4. Different systems
5. Different prices
6. Different geography
7. Different transportation competition
8. Different cultures
9. Different densities

But of all those, you think #9 is the reason it's doing "better"?

Who's to say that if you spread it out over a wider area it wouldn't get 10 trips per bike?

If you want to convince me, find the density, tpbpd, and population denstity of every system in the world at 6 month intervals. They we can plot out the data and try to come up with some conclusions. But otherwise this is apples to oranges and borderline irrelevant.

Washcycle, you're right, we could run a regression with every bike share system, input the station density, population density and all the other variables, and get an answer to which one has the highest correlation. I'm not about to do that though, although anyone who is reading this is welcome to try. Might make a good college thesis.

Also, how much denser is mexico city? Like DC, it's mostly low buildings, because of the earthquakes. Yes, reforma has the tallest building in latin america, but go one block back, and everything is 5 floors high at most.

However, I think it's easy to agree that a cleachannel closed system (annual members only) is a huge point against usage right? I mean, look at smartbike, look at rome, look at milan....look at most clearchannel systems.

Basically, I've personally visited and looked up details for the following bikeshare systems:

DC, old and new
Baltimore (proposed/cancelled)
Miami Beach
Mexico City

Some are pretty big failures. Some are massive successes. From what Ive gathered, the three most important factors are:

-Helmet requirement (Instant fail, see australia)
-Subscription requirement (see most clearchannel systems)
-Density and number of stations (see smartbike, rio)
-Density of city (ie, won't ever work in Phoenix, would be very successful in downtown Pittsburgh)

On the other hand...
-Weather? Not really a factor. That is, it won't work in Siberia or the Sahara....but we dont build cities in places like that for a reason. Most major cities have year-round mild temperatures. And yes, 25f-95f is mild.
-Culture? Not really a factor.
-Infrastructure? Not really a factor. Paris has no bike infrastructure at launch.
-Traffic? not really a factor. Again, look at Paris and Mexico city.

Again, this is without running the regression to see what the numbers say, just my own observations.

So in summary, because it would be politically difficult to yank out existing stations (the ones I mentioned before) to relocate them, DC must ensure that the next 50 stations fill in the existing service area (downtown, eotr, the central universities). No random stations in Takoma.

If Silver Spring/Takoma want bikes in the next 2-3 years, wonderful, it needs to be set up like an independent system, like Crystal City. 25 stations, minimum. People wont ride from Takoma to K street. They will ride from Takoma to the SS movie theater, and SS residents will ride to the Takoma recreation center (and pretend they pay taxes in DC).

@Mr Share - it's a minor point, but there are some rather large cities in Siberia. Novosibirsk has 1.5 million people, and Irkutsk has ~500,000.

If you take Mexico City's system and overlay it over DC you get a system running between the Mall and Mass Ave from 19th Street NW to Union Station and then south to the baseball stadium. I don't think that system would do as well as the one we have. I wouldn't join that system.

As for calculating car trips removed, that was done in Montreal using surveys, so that can be done. I feel like Mexico City's is mostly replacing walking trips and taxi rides. Not so bad, but not really what we're going for here.

Also, how much denser is mexico city?

About 5 times denser.

Weather? Not really a factor.

What? Montreal shuts theirs down in the winter. In DC, do you really think ridership will be the same in January as it is in May? I think they're talking about a 30% drop off during the winter months. A four month drop off by 30% means DC will need to average 10-11 tpbpd in the other months to match Mexico City's 9 for the whole year.

Culture? Not really a factor.

I think the places it's been tried are so similar that it's hard to say that. But a bad "bike culture" has been given as a reason for poor performance at times.

Infrastructure? Not really a factor. Paris has no bike infrastructure at launch.

Absolutely not true. There are bike/bus lanes and streets closed to cars and more as I recall.

washcycle, you're right, overlaying it like that wouldnt be as successful. Which is why the system should have launched bigger. Smaller geographic area with 50% more stations.

Smartbike was a prime example of "if you cant do it right, dont do it at all".

Obviously Cabi hasnt been a disaster, but if they dont keep adding stations and filling in their holes, they will lose members who are tired of dealing with the problems caused by the geographic sparsity of the stations.

By weather, I mean bike share works in cities as varied as montreal, miami, milan, oslo and mexico city. Saying "it gets over 80f here, no one will ride" and "it gets below 40f here, no one will ride" si simply untrue. If it's snowing, less people will ride...but people still will. I think Montreal is making a mistake by shutting down in winter, as it says "bikes are only useful on fair weather days"

Yes, Paris has bike lanes now. They didnt in 2007. The bus lanes were unofficial bike lanes, now they added signs. Other cities in France were much friendlier to bikes. Le Havre, for example, has an Amsterdam-style network of separated paths. Bike share in paris has launched many changes in how they distribute streets. Hopefully the same happens in Rome which also has garbage bike infrastructure. (The north of Italy is much better). Washington is already reaping some of those rewards, as even the examiner has grown friendlier to infrastructure because of all the new riders.

People used to say "bike share wont work if you dont put the lanes down first" and I always argued it was the opposite. I still believe I'm right. Build a bike share system, and the bike lanes will follow.

The comments to this entry are closed.

Banner design by creativecouchdesigns.com

City Paper's Best Local Bike Blog 2009


 Subscribe in a reader