A look at Hubway in Cambridge, MA
MIT, millennials, and the weather shape bike sharing in Cambridge
Bike share programs have popped up in cities across the world at a rapid pace. They are often touted as having many benefits including reducing pollution and congestion and improving public health. Despite varying success across cities, there’s seems little doubt that their popularity isn’t on the rise. This is evident in the rise of even more bike sharing models like dockless bike sharing.
Another growing trend in cities across the world is the use of data in governance. Just as businesses and sports teams have jumped on the data train, governments are beginning to as well. In the near future, all cities will govern by using data intelligently to better serve their residents.
In the interest of exploring these two trends, I wanted to look at how bike sharing was being used in the city I live in, Cambridge, Massachusetts. Most of the existing analyses done on the Hubway bike sharing program either look at the program as a whole or focus on our bigger sister city to the south, Boston. So I thought it would be interesting to just look at how Hubway was being used in Cambridge.
Without further ado, let’s dive into what I found.
Commuting with Hubway in Cambridge
The first interesting, though perhaps not that surprising, finding was how central MIT is to the Hubway program in Cambridge. Out of the top ten most popular trips in Cambridge, only trips between Linear Park and Davis Square did not involve a station at MIT.
This is perhaps not surprising given that MIT is near Boston, a more popular city, shown by the fact that the most popular trips are between Beacon St in Boston and the MIT Mass Ave station. However, it does reveal how central MIT is to Cambridge. Perhaps more surprising is that no Harvard stations appeared in the list of the most popular trips taken.
This leads to the question of what is Hubway mostly being used for in Cambridge? Are Hubway rides typically made by tourists? Students? Professionals?
So I looked at when trips were being made and found that there were more trips taken on a typical weekday than there were on a weekend day.
This was a bit surprising to me. I thought that Hubway might have more of a casual, take a nice Sunday ride, kind of usage but in fact it appears to be the opposite.
I then looked at what time of day most rides were taking place and found what you would expect if Hubway was being used as a way to get to work. Looking at trips taken during the week, the most popular time of day to take a trip was between 4pm and 8pm, with over 30% of all rides, and between 5am and 10am, with over 25% of all rides. So more than half of all rides during the week occur during the typical work rush hours. This is actually even understating its usage for commuting to work since the science and engineering crowd aren’t always up and at them first thing in the morning, which would make the 18% of rides between 10am and 1pm to also fall into this category.
Hubway seems to be being used as an alternative means of public transportation for people getting to and from work.
It also appears that a lot of Cambridge rides are short. 80% of trips happen under 20 minutes. This isn’t that surprising given that Cambridge is not all that big of city to begin with. But it does indicate that most of the commuting done on Hubway probably does not replace other means of transportation (like driving, buses, or the T) and instead fits as an alternative to walking long distances or taking short bus or T trips. This would counter the argument that it reduces pollution that much if it is only replacing trips that otherwise would have little impact on pollution. This requires further investigation.
Who’s Hubwaying in Cambridge?
Another thing that I found interesting while looking into the Cambridge Hubway data was the demographics of the typical Cambridge Hubway rider. I was surprised to find how much more popular Hubway is with those aged 25–34 than with those 16–24. Thanks, millennials.
Based on this information and the information above, it’s pretty obvious that in Cambridge, Hubway is mostly being used as a mode of transportation for the young professional. Kendall Square is home to many companies’ offices so it makes sense that a lot of people would be commuting to that area during the workday. It’s also possible that graduate students make up some of these numbers since they tend to live off campus but given that there are only about 7,000 graduate students at MIT, they probably don’t make up that much.
The other surprising piece of information found about age of riders was that the 45–64 year old age range holds its own against the younger groups. This most likely reflects Cambridge’s bike culture but is impressive nonetheless!
One question that I have of in regards to this data is whether or not the ridership by age will always tend to be young because that’s the range when you are less likely to have a family, a car, and live in the suburbs? Or is this a generational change and our generation will continue to use options like bike sharing as a means of transportation even as we get older? I suspect it is the former but only time will tell.
Finally, another interesting finding on the demographics of Hubway riders in Cambridge was the gender differences. Males made up over 60% of all rides in Cambridge and possibly more since there was a decently large number of unknowns or others reported (this is self reported according to Hubway).
It’s possible the difference in ridership amongst genders is due to clothing though it could be for other reasons as well. This is one area that should be further explored.
It’s always about the weather in New England
One final interesting piece of data that I came across in my exploration was the impact of weather. Being a non-native of New England this one is particularly not surprising. Despite what you may here from die-hard New Englanders, the weather here is in one word, terrible. It’s also apparently terrible for Hubway too.
The below graph is all of Cambridge Hubway trips per day over time. As you can see, there is a lot of variability. Even if you ignore the significant drop-offs for the winter (a particularly brutal time to be riding a bike in Cambridge), you’ll notice that in the summer months there are still a number of drop-offs. I explored this data further and found that they almost always correlated with poor weather. Sometimes that had to do with unusually cold weather and sometimes because of rain. Of the two, rain had the larger impact and largely explains the biggest drop-offs found in the graph below.
Weather is known to have a big impact on bike sharing programs. Kaggle, the data-science competition website, has even hosted a competition for predicting ridership based on the weather. But it is still interesting to actually see the impact in the data and how significant particularly bad weather days can have. This also leads one to consider how cities could mitigate weather’s effects on bike sharing by investing in infrastructure.
It was pretty interesting going through the data and looking at how bike sharing is used in the city I live in. It’s not only revealing about how Hubway is used by Cambridge residents but also how Cambridge residents go about their daily lives in general.
There is obviously a lot more you can do with this data. As a resident of Cambridge, I hope that the city looks at data like this to make better informed decisions about how to serve its residents. For example, if the city was interested in increasing bike sharing amongst its residents, it should consider improving the biking infrastructure around the city.
This could be done in a number of different ways. They could increase the amount of people across age ranges using bike sharing by improving the safety of biking in the city with more protected bike lanes.
They could also add new infrastructure to improve the biking experience in bad weather. One idea would be to install bike lane shelters above the most popular bike routes across the city. They could even fit these with solar panels to have a dual purpose and reduce overall cost in the long term.
Bike sharing is just one city program that can use data to better improve the lives of city residents. I hope this exploration into how it is used in Cambridge is informative and inspires new ideas for how city services can be improved.
Unfortunately, I didn’t explore everything that I thought of but here is a list of things that I think could be further explored in relation to this post.
- Look at how different variables, like weather, make bike sharing trips predictable. The future of cities and data will be proactive rather than reactive. With this in mind, how predictable is bike sharing?
- In 2015, Cambridge Hubway trips dropped off, why?
- What policies, infrastructure, etc can a city implement to encourage more use of bike sharing programs?
- Look at how some of these stats have changed over time, identifying trends
- Dec ’16 data is missing, would be good to have it to see if there is any new discoveries
- Why women are so outnumbered? Is it because of clothing? Something else?
- Keep it going for 2018 data
- Create a real-time version of this analysis with Hubway’s real-time data.
- Create interactive maps detailing all the above information, I’ve kind of started on this but unfortunately Medium doesn’t support embedding Mapbox maps.
One of the benefits of having public bike sharing programs is that their data is made available to the public. You can find all of the data I used for this analysis on Hubway’s website.
Also, in the interest of replication and exploring the data yourself, you can find all of the code and data that I used in this post here. Please let me know if I made any mistakes or am missing something, I welcome the feedback.
Enjoy this analysis? It’s part of a series of data analyses I am doing on the city I live in, Cambridge, Massachusetts. You can find the rest of them here: https://medium.com/a-look-at-cambridge.
 I define Cambridge Hubway usage as any Hubway trip that either started or ended in Cambridge.
 Based on my own experience as an engineer ;)