Capital Bikeshare Data, Part 3

A short Part 3 in a series of posts analyzing Capital Bikeshare usage data. This post quickly looks at how much each bike gets used and churned through the system. See parts 1, 2, and 4.

When you get on a Bikeshare bike, what’s that bike’s history? Has it been all over town in recent weeks, or are you the first to use it in awhile? How evenly is each bike in the fleet used? Do some bikes do all the work, while others sit?

The dataset of all trips counts 1,343 unique bicycle ID numbers. I’ve excluded 25 bikes that have strange ID numbers (they look like “? (0xEB7B5641)” if you’re interested), and suspicious looking activity, but they only represent 0.3% of all 1.36 million trips.  Here’s how the remaining fleet of 1,318 bikes breaks down by number of trips over 2011, excluding the 2010 growing pains. It looks like Bikeshare introduced about 200 new bikes to the fleet in late October 2011 (the ones with the flashing LED front lights?), so I’ve called those out separately.

Capital Bikeshare Fleet by Number of Trips. Most bikes were used 1,000 to 1,500 times in 2011, but some were used much less.

So in 2011, most bikes in the fleet were ridden between 1,000 and 1,500 times. No bike was ridden more than 1,600 times. Interestingly, a bunch of bikes were used at half, or even a quarter as much as the average bike. This could’ve been because the bike was pulled from rotation in the winter when the fleet is reduced, or for maintenance reasons.  I can’t tell when a bike was pulled for maintenance, but I can exclude winter.  Here’s all trips in 2011, but from April to October only when the fleet was presumably in near-maximum service (new fleet not shown here):

Capital Bikeshare Fleet by Number of Trips, High Season Only.

All the numbers come down since we’re only looking at the high season, but still a fair chunk of the fleet seems to be getting used alot less than average. What’s going on here?

Are bikes getting stuck in less popular stations?  How completely does the system churn? Are there pockets of bikes that get stuck on the back bench for awhile, only to hit the big time weeks later?

One way to look at this is to say – Of the bikes with only 300-800 rides in 2011, were there lulls in usage? What station were they last docked at before a lull, and is there a pattern in those stations? Do those stations correspond to the usage numbers by station?  I’ll need more time to set up that query – anyone want to take a crack at it?

One parting thought – what percent of a Capital Bikeshare’s life is spent sitting at a dock?  I get something like 415,000 hours of total usage in 2011 (bike-hours, if you will).  Each year has 8,760 hours in it, times say 800 bikes in service, and that’s 7 million bike-hours of total availability.  So each bike is being ridden 6 or 7% of its waking (and sleeping) hours, while the other 90+% of its life is spent on the dock? Is that low or high? How does that compare to a typical car, or bike?