COVID-19 has drastically increased the scale, speed and ambition of many of the world's cities to become more cycle and walking friendly. Milan for example has brought forward its 2030 target to convert 35 KM of city road into cycle and walking lanes forward to 2020 and London is turning central parts of the city into one of the largest car-free zones in the world.
Various forms of crowdsourcing and citizen-generated data can help planners make sure these changes meet the needs of citizens.
Every journey we make can provide vital data for anyone thinking about where and how to design new cycling and walking infrastructure.
When aggregated, our individual choices on everything ranging from whether we take the easiest or least polluted route or how we find shortcuts to get from A to B as fast as possible provide a detailed understanding of how we use existing infrastructure and what is missing.
Strava Metro is one example of this. Using data generated from more than 40 million users worldwide who track their cycling and running through the Strava app, the company helps local authorities understand how people move around the city. In London anonymised data from the app, has helped inform decisions on planning for the city’s cycle superhighways. This data can also be used to understand the impact of cycling infrastructure investments i.e whether more people who are using the app are cycling on routes with new cycle lanes. Similarly, crowdsourcing ideas from the public can help cities make more informed or targeted decisions. The See Sense app for example crowdsources information from bicyclists about dangerous roads.
Collective intelligence can also help people plan the routes that best fit their needs. Companies such as Beeline are providing crowdsourced information of better and safer cycling routes. The GoJauntly app lets people access green city walking routes as well as create and share their own and has partnered with TFL to improve on their suggested walking routes between Tube stations.
Focusing on the same challenge, researchers, using crowdsourced data from the online game Scenic-Or-Not have explored whether algorithms can be trained to identify what humans consider ‘scenic’. This could help add features to journey planners and enable people to plan routes that prioritise a ‘scenic’ experience over speed.
Others initiatives have focused on how to help people avoid unnecessary pollution through visualising levels of pollution, often through using a combination of data from city sensors and crowdsourced data from people travelling around the city. Recently one of these platforms, Plume Labs, announced that they would be able to provide users with street by street air pollution information to help them plan their journeys.
Crowdsourcing accessibility issues
Making cities better for cycling and walking isn’t just about the number of cycle lanes and pedestrianised space, but also ensuring that everyone is able to use them.
Getting around most cities remains a challenge for people living with a disability. The average city street often includes sandwich boards that block your path, uneven or steep kerbs on pavements which can all hinder one’s ability to navigate the city freely.
Crowdsourcing plays a big role in mapping accessibility issues in cities. Perhaps the most widely used tool for this is wheelmap.org which, using Open Street Map, lets users map spaces that are or aren’t wheelchair accessible in cities all over the world.
Another example of a crowd-based approach to mapping accessibility issues is US based Project Sidewalk. The project looks to address the lack of data about accessibility of existing infrastructure to people with mobility impairments through the use of online crowdworkers to remotely label pedestrian related accessibility problems by virtually walking through city streets in Google Street View. Crowdworkers and volunteers begin their journey with a simple-to-use online tutorial on what to look for and how to rate sidewalk accessibility, such as missing curb ramps, sidewalk obstacles, surface problems, and the lack of a sidewalk on a pedestrian pathway. Following the tutorial, users are assigned missions to complete a certain distance of sidewalk segments in the city or choose a specific neighbourhood.
The model is showing real promise: a data validation found that, on average, users could find 63% of accessibility issues at 71% precision,
Peter Baeck is head of the Centre for Collective Intelligence Design at Nesta and Sophie Reynolds is a researcher at Nesta
