As a visitor to Oslo, you will quickly find a robust bike sharing culture with city bikes being one of the most popular forms of public transport.
The shared bikes tend to be available where you need them, when you need them. If your curiosity leads you to dig into the stats behind this, the number of trips per bike per day in Oslo is eight. This is four times the global standard, and one of the best shared-bike usage rates in the world.
Norway’s second and third largest cities after Oslo only received city-wide bike sharing schemes last summer, but their usage rate is predicted to reach the same impressive average as in the capital. But what is it about the Norwegians that makes them embrace bike-sharing so wholeheartedly?
There are a few possible explanations. First, Norway has for many years had a strong public transport system. Bike-sharing tends to be viewed as a natural part of how the people in the city move.
The next reason is, of course, tech. People in Norway grew up surrounded by digital solutions, and so are accustomed to relying on, and trusting, tech. There is also strong network connectivity throughout the majority of Norway, which of course empowers digital cycle-sharing schemes. And this is even before 5G takes over.
However, one might argue that these characteristics can be applied to most Scandinavian countries. And yet Norway’s bike-sharing schemes currently outperforms those in both Denmark and Sweden. To explain why that is, we look to data.
In 2016 we realised we were sitting on a trove of data, and that this could help us understand the actual needs of our users in Oslo.
The first thing we found was that, while people were allowed to rent bikes for periods as long as three hours, they normally only used their bikes for a fraction of that time. There were long periods of time when a bike would be unavailable, often for no good reason.
For the next season we cut the max usage time down to 45 minutes. The vast majority of users felt no negative impact by this change, but it did lead to a drastic increase in the number of bikes available at any one time. Just reducing the rental time completely transformed our system.
Further analysis showed that not every location in Oslo had the same demand for bikes, nor was the demand consistent throughout the day. Thanks to our team of data scientists, we are able to correct the system in real-time, through an intuitive rebalancing model and people on the ground. Data analysis allows us to immediately address concerns and accommodate changes in user attitude and behaviour.
With intelligent data analysis, cities can get a better idea of the “how” and “where” behind cycling in an urban metropolis. We’ve found this is also true for the “who” in a study on gender we conducted with TØI, the Norwegian institute of transport economic.
Here’s what we found: In 2017, 32% of all our trips were taken by women, even though about 41% of our users are women. More women than men said they would like to see more designated bike paths in the city, and more women than men said they would use bikes more if those paths existed. There was a strong connection between areas in Oslo which featured a large number of women employees, and the majority of biking routes taken by women.
When we paired this data with other insights, it seemed clear that the distribution of stations and rebalancing of bikes inherited a bias already existing in the city. We’ve adjusted our scheme accordingly.
And we don’t keep this data to ourselves. Ruter actively uses our data for its own logistical planning. Entur includes the number of available bikes in its app. Students, researchers, and enthusiasts are all welcome to analyze our data as they like.
What can other cities takeaway from our experiences? First, data is key. A city needs to give its population the micro-mobility solution that they want, in a way that scales to the unique needs of that city. It can’t do that unless it knows how and where the users make use of a micro-mobility scheme.
Second, standard rebalancing algorithms do not intuitively account for gender or wealth distribution. If a city wants to make sure each population is served equally, they must study and accommodate demographic usage patterns isolated from average usage patterns.
Urban Sharing began as a startup focused on the bike-sharing scheme in Oslo. Now, we’re looking to find ways to apply our existing models to other cities.
We’re working with a team in Trondheim to identify how to use machine-learning to help adjust our model to the needs of cities outside of Norway. This will get better the more cities we have and help us to work towards establishing a global sustainable solution to the challenges of micro mobility.