In 2013, the city of Nice, France, launched an ambitious “smart parking” service that allowed drivers to find available parking spaces in real-time through a smartphone app.
In order for the system to work, the city paid 10 million euros (or about $11.3 million USD) to install hundreds of sensors next to parking spaces, as well as 291 payment booths across the city. This was a steep price tag, but the system quickly became a model for innovation within smart-city circles and gatherings.
However, Le Monde reported this May that the city has shut down the system and is returning to old-fashioned parking meters. How did a smart-city case study turn into a 10-million-euro failure? The same article suggests that its usage wasn’t properly tested. For instance, authorities failed to consider that people shouldn’t look at their smartphones while driving, or that double-parking is rampant in Nice and reduces demand for the designated spaces.
Whatever the actual reasons were, this 10-million-euro flop raises an important question: How can we anticipate how people will ultimately use innovative urban projects?
Lessons from Calgary’s quick-build bike tracks
The obvious answer is that cities should apply user-centered design principles when building these sorts of projects (such as simple test-and-learn methods). Unfortunately, the politics involved in urban projects have a way of making rapid prototyping not so rapid, which means that cities need to find creative ways to gather insights from users.
A great lesson can be learned from Calgary’s mayor Naheed Nenshi. At this year’s Bits and Bricks forum, Nenshi explained how his city tackled both the complexity of planning infrastructure that will actually be useful to citizens, and the usual public resistance to allocate car space to bicycle lanes.
In 2015, the city deployed a one-year pilot program with temporary quick-build bike lanes. This means that test tracks could be deployed fast, not only from a construction point of view, but also politically. Indeed, the mayor pledged to keep a close watch on the consequences of the lanes and adapt them if necessary:
“I believe in experiments and I believe in trying things out, so we’re going to measure this very carefully and measure the impact on cars, we’ll measure how many cyclists there are, we’ll measure the impact on local businesses.”
To achieve this result, the city equipped the lanes with sensors at strategic points in order to measure bike usage. This allowed it to confirm usage patterns that would otherwise be considered assumptions and, by making this data accessible to anyone online, it eased political consensus by making the whole experiment fully transparent.
According to Nenshi, despite the fact that bike traffic tripled along specific lanes in the first month, some skeptics would still argue that the tracks weren’t being used. Not anymore: “My response is ‘go to this website and you will see precisely how many bikes went by this track on that day,’” said Nenshi.
When data-driven design meets transit in Paris
The case of Calgary’s quick-build lanes can prove useful for future projects where there is no impeding infrastructure in place. But how do we gain access to similar insights when working around established large-scale infrastructure? For instance, how would we achieve the same results for the massive suburban rail network in Paris without spending millions in gathering data across the region’s 500 stations?
France’s railway operator SNCF explored this idea during a recent research study. The idea was to link data from trip-planning websites with real on-site statistics such as ticket validation and selling data.
By analyzing 100 million requests made across three months, SNCF found a very precise correlation between online trip-planning requests and the other data sources that indicate passenger traffic. These findings could prove to be extremely useful, not only because it is a very cost-effective way to gather data, but also because it allows the anticipation of traffic from between a few hours to a few days in advance.
Transportation operators can use this data to adjust services in real-time when there is an anticipated traffic spike (for example, allocating more people to guide travelers at train stations) or it can serve as data for studies and transport planning when traditional sources are nonexistent.
As Nenshi says, good data drives good decisions.
Photo, top: Crowds at Paris-Gare de Lyon train station (Patrick Janicek, Flickr, Creative Commons).