Our Mobility in Knots

Our Mobility in Knots

Thanks to a study conducted by MaaS L.A.B.S. at the University of Applied Sciences Potsdam and the University of Siegen, we were able to work with a dataset of 32 participants. They mostly are residents of a small part of northern Potsdam. Not only did they record their GPS pathways. They also collected the mode of transportation and the reason for each, specific trip. That gave us a unique insight into a set of very privacy-sensitive yet highly interesting data points that possibly tell us more about how individual mobility works.

This data was recorded in early 2021, right when the pandemic was peaking again and certain lockdown measures were in place. That's why our dataset is also a glimpse into the world of mobility during a global pandemic.

We have selected four different people for the visualization. Of course, there are many more participants in the dataset. We carefully considered who to select and stuck with the people who had at least some mobility despite the lockdown. From these individuals, we selected four people with distinct mobility patterns. That means different main modes of transportation, times, routes, and regularities within mobility. Even though this is of course anything but representative, we want to use it to show you a diverse spectrum of mobility patterns.

We asked ourselves how we can create a meaningful, rich visualization of the data while still ensuring the participants' privacy - and came up with mobility networks. But before we take a deep dive into our data, let's first go for a trial run...

This is the simplest possible mobility, as it only consists of one mode of mobility: Walking.
Just follow the arrow that begins at the knot “Start”. That will lead you to the mobility mode the person has chosen. From this point, you will get to the end of the path.
As we can see, this person is starting and ending walks - nothing more.
The more a participant used a specific switch between mobility modes, the thicker the connection between the knots gets.
There is no information about the time or the place of the mobility in the network. That way, we ensure the privacy of each participant while still establishing a meaningful visualization. To give you some reference on the impact of each mode of transport, we added a small chart on the lower left. There, you can see how much time and how far this participant traveled with each mode in one trip on average.
Let's make it a little more complicated and see what happens. This person walks and has a car, so two different modes of mobility. Sometimes the person uses both, maybe taking a walk after a long drive or simply having to walk to the parking station.
Driving a car is closely linked to walking in this example - and wouldn't be possible without it. The explanation is straightforward: a person must walk to get to the car.
This close link will get more interesting later when we see how public transportation and individual supporting mobility accounts for a flexible model of mobility that allows being spontaneous.
This is the first graph representing the mobility of one of the participants during the lockdown of march 2021. We can't see how the participant's mobility was before the pandemic, but we can clearly identify a focus on individual car-centric mobility.
It is the mobility of a person working nine-to-five. Getting to and from work by car and occasional errands at the local supermarket are the main reasons for mobility.
Here are some quotes from the conducted survey: "For me, my car is a necessary commodity that should only serve its purpose."
"Due to the pandemic, I use public transport rather less and the car rather more often."
"The car is my main means of transport. I think my mobility is typical for where I live."
When moving away from classic car-centric mobility, things can get complicated fast.
The participant stated in the survey: "Bicycling is my main mode of transportation and I do not own a car."
and "for me, avoiding rush hour was the biggest challenge to adapting my mobility behavior to the Corona pandemic - driving less to the office was not a challenge."
Also, the participant has some wishes for their neighborhood: "there is a lack of a healthy food store, public playgrounds, swimming pool, recycling center, and vegetarian restaurant in my neighborhood."
Their mobility seems to be massively influenced by the pandemic. Most days, this person barely leaves the house - only for an occasional walk during the day.
Other days, the home office doesn't seem to do it. The participant travels by bike to the Main Station, waits for the regional train, and then heads off to work. On their way back, they run an errand or…
...pick someone up, maybe their kid. For this participant, the reason for using the bicycle is mainly the shorter travel time. This contrasts common beliefs, that this mobility is more time-consuming because it could depend on fixed schedules of trains or the physical place you locked your Bicycle.
If you look at the visualisation, you can see a strong relationship between bicycling and short stops and vice versa. This could be related to the fact that traffic is not optimized for bicycles, so they have to stop frequently. Even though biking is the fastest mode of transport for the participant, it could still be better.
The next participant has a diverse mobility and uses all modes of transport for different occasions.
Despite having a job ticket, the person drives to work by car. One reason for this could be the pandemic, as the person uses public transport much less as a result.
Oftentimes, the persons combines the car ride to work with other errands. For example, they bring their kid to kindergarden or get groceries.
In their free time, the person uses different means of transport such as the bicycle or public transport. They like to go for a walk in the nearby forest. For longer distances such as to the southwest of Berlin and shopping the car is used. Most mornings, they bring their kid to kindergarden with a bike.
They say: "I bought the car for routes that are not accessible by public transport (last mile)."
"My neighborhood lacks a pharmacy, a youth center, a swimming pool, a sports club with a sports field, or a sports hall."
The nearest bus stop is less than a kilometer from home and very accessible.
"Due to the pandemic, my mobility has decreased considerably, but it does not pose any major challenges. If anything, my mobility has become more climate-friendly. I use the bicycle just as often, but public transport and the car much less. Nevertheless, the car is my main means of transport."
Similarly to our first person, the next participant travels from Potsdam to Berlin for work.
For transit, they use public transport. It takes them about 50 minutes to get there.
Taking the car would take 15 minutes less. But traffic and the time to look for a parking station is not included, so it would probably not be that much faster.
To further cut down the transit, they occasionally take their bike with them on the train and use it to get from their home to the first train station. Sometimes, they use a bike-sharing service, but less often since the pandemic.
Even though they normally optimize their transit time, sometimes it does not work out: "Today I missed the connecting regional train." The walk between the two stations is too long. They needed to wait 20 minutes for the next train to come.
They own a car in their household. About that, they say: 'We did not buy it, we just borrowed it from a family member as they don't use it at the moment.'
They do not give the car any emotional value or memories. It only serves its purpose, but they do not need it specifically.
At their neighborhood and their workplace, they wish for roofed parking slots for the bike
Introduction
1.6 km
startwalkend

How far and how long do the participants on average travel with each transport mode?

The networks visualize individual mobility patterns while allowing a look into relationships and links between modes of mobility. As you could already tell from the small graphs on the lower left, it is nonetheless important to reflect and compare the impact of each mode of transport. Therefore, we now look at all participants of the study together and visualize, how far and how long they travel on average with each mode.

1.85 km19.65 km3.24 km14.67 km

How did we calculate this averages? Each of the recorded trips consists of smaller legs with different means of transport that the person needs to get to the destination. So, to calculate the average length and average time for each mode of transport, we had to be careful that the small legs did not distort the result. Therefore, we first calculated the total distances for each mode of transportation within a trip - so it doesn't matter if you had to change trains during the trip. From these total distances, we then calculated the average.

As expected, people travel on average with the car and the public transport the farest. The difference between the two is only a few kilometers. The networks you saw before hihlighted the more frequent use of bicycles and walking. However, this bar chart shows that in total one single trip with the car or public transport, respectively, has a higher impact.

So what is our conclusion in total?

Mobility is individual

Each of us has his or her personal mobility preferences. How do I move around in everyday life, perhaps to experience beautiful moments in nature in between? Which shops do I integrate into my routines because they might have the best bread rolls in town? Where do I have to be at what time, and therefore can't spare a minute? These decisions and connections are highly individual and often difficult to capture. However, with this survey data, we were able to gain insight into individual mobility and the unique motivations of the participants.

Mobility is sensitive

Every start point and destination, every transferring time, and every change of means of transport reveal something about the person. Working with this kind of data made us more aware of what can possibly be done with all those data points and also how we can visualize them without attacking the privacy of participants in the study. That's why we choose to look at networks and the interconnectedness of means of mobility, supported by average values of time and space.

Mobility is plural

Our mobility behavior responds to various factors: the range of different means of transport, the availability and quality of public transport, our time budget, and, as can be seen well in the data, the current (pandemic) world situation. Mobility, however, never remains one-dimensional: Being mobile in a city means moving between modes of movement, looking for a combination of modes that work well together and suit your individual needs. These connections of means of transport give us reason to take a close look: which transfer feels good, which waiting time annoying? How many rental bicycles are enough when everyone has to go to work in the morning?

Credits

Helmut Büttner

Research, Data Analytics & Narrative Structure

Johanna Hartmann

Conception, Data Analytics / Processing, Data Viz & Web Development

Sascha Höver

Design, Conception & Co-Design Workshop

We finished the project on 28th February 2021.

We developed this project for the course Mapping Cities – Making Cities by Prof. Dr. Marian Dörk at University of Applied Sciences Potsdam, Winter semester 2021/22. Thanks for your support and feedback!

Special thanks also go to Christian Berkes and the MaaS L.a.b.s. team for the collaboration, your feedback and for providing data for this project.

Study

Find out more about the MaaS L.A.B.S project on their website and keep up to date with new reseach and events: "https://www.maas4.de". The invitation to participate in this study "digital traces potsdam" is here: https://www.maas4.de/movinglab

Stevens, G.; Pakusch, C.; Böhm, L.; Bossauer, P. (09.09.2021). Digital Traces Potsdam. DLR MovingLab Studie. Universität Siegen, MaaS L.A.B.S., https://movinglab.dlr.de/en/projects-campaigns/digital-traces-potsdam-maas-labs-research-project

References

Dörk, M. (18.06.2020). Experiments in Temporal & Relational Data Visualization [Slides]. University of Applied Sciences Potsdam.

Horne, Rp.; Jeffrey-Wilensky J.; Kranz, M.; Ozelli, K.; Stefaner, M. (12.06.2020). Reseacher Connections: Understanding a decade of collaborations in autism science [Project]. Spectrum News. https://connections.spectrumnews.org

Otten, H.; Hildebrandt, L.; Nagel, T.; Dörk, M.; Müller B. (2014 - 2018). Shifted Maps: Revealing networks in personal movement data [Project]. UCLAB, University of Applied Sciences Potsdam. https://uclab.fh-potsdam.de/projects/shifted-maps/

Sedlmair, M., Meyer, M., and Munzner, T. (2012). Design study methodology: Reflections from the trenches and the stacks. IEEE Transactions on Visualization and Computer Graphics, 18(12):2431–2440. https://doi.org/10.1109/TVCG.2012.213

Meyer, M. and Dykes, J. (2019). Criteria for rigor in visualization design study. IEEE transactions on visualization and computer graphics, 26(1):87–97. https://doi.org/10.1109/TVCG.2019.2934539