What makes a city smart is having the right information at the right moment for the right stakeholder. Opening up city government data is the first step to that, and some cities are actively doing so right now – New York City, San Francisco, Boston in the U.S., for example. Those that aren’t should perhaps take a peek at the Urbanopoly mobile app for Android devices, which takes the approach of consuming, creating and assessing the quality of Smart City-related data.
“I believe that this kind of application could convince people and public authorities to release their data set even if the data is not high quality,” says Irene Celino, senior researcher and senior consultant at CEFRIEL which, with the Politecnico di Milano within the PlanetData Research Project, developed the mobile game based on human computation and Linked Data technologies. The effort was co-funded by the European Commission under the 7th Framework Programme. “Even if they’re not sure of the quality of the data you can still make it open and ask the community to improve it. This is the moment in which a real collaborative system can be built between individual citizens and administrations and app developers and scientists.”
Introduced in May and debuting on Google’s AppStore in July – and presented last week at ISWC — Urbanopoly, one of the finalists of the Semantic Web Challenge 2012, is a social, mobile and location-based game with a purpose that started with the Milano region in Italy and then added Amsterdam and Boston locations.
That purpose is to collect and verify and cross-check geospatial data, starting with the Linked GeoData (which uses the information collected by the OpenStreetMap project and makes it available as an RDF knowledge base according to the Linked Data principles) set. Players with their GPS-enabled, location-positioning mobile devices become citizen sensors, providing what they see and experience – but they are not just collecting data, they also are validating it.
“We not only want to collect information we don’t know, but we also want people to validate pre- existing data,” says Celino. Players check in at different locations on the map around them and spin a wheel of fortune that gives them a challenge to pursue, such as providing data by being quizzed about specific features of a venue – like a business or public transportation or monuments. If they’re provided with pre-existing values related to a venue, they can check if it is the same as previously described or not. With OpenStreetMap, Celino explains, there can be a lot of variety in the information that’s provided, some of it very detailed and some of it very basic or outdated – a business might have closed, for instance, and been replaced by a new one.
“The idea is …geographic information systems suffer from the problem of continuity, of provisioning of data or checking or verification. This game doesn’t solve all the problems, but it is a medium to provide an additional channel,” she says. And it’s built to entice not just those active contributor types who are naturally proactive about efforts like OpenStreetMap, but also the person who just wants to play a game, where data collection is just a side effect of their fun. Urbanopoly is publishing its output as Linked Open Data, too. “We built a full game with a lot of gaming features that has nothing to do with purpose, but are entertaining,” she says, such as players getting ‘rents’ for the properties they come to own and publishing a leader board to keep the competitive juices flowing.
So far, Celino says the game is working to collect data that was missing before but that probably the main improvement is helping with detecting changes. The game processes all the data and uses an algorithm to compare the information that is imparted, to assess answer accuracy from multiple sources – and to discourage cheating by providing inaccurate data, the game penalizes users, too. What perhaps is a bit more of an issue is that there are so many venues in a city like Milan, so it’s hard to get scale for any one place. “You can get to consolidated information when more people play in the same places,” she says. “There are more than 50,000 venues in the system and that makes it difficult for now around 100 users to do cross- checking on the same data.”
The Average Life Play (ALP), or how long users are engaged in the game, is very good, she says, at about the 100 minute mark and continually increasing. The other metric is number of tasks solved in a unit of time, and Celino’s assessment is that potentially people have a very high rate of contributions. That’s good news for the project as is, but also when it comes to the idea of turning a game with a purpose to new ends, such as helping cities more efficiently assess origin-destination matrixes.
Governments spend a lot of money on those surveys, she syas, so why not think of a game where people check in when they get on a bus and check out when they get off, or arrive at home or the office? “It’s giving a different way, a different incentive to collect the same kind of data,” she says. “These are challenges in the real world.”