Tom Erickson, designer and researcher in the Social Computing Group  at IBM's Watson Labs in New York, believes in crowdsourcing as a force capable to bring to citizens the ability to change the landscape of their cities and effectively create a more intelligent and livable environment. In his paper Geocentric Crowdsourcing and Smarter Cities: Enabling Urban Intelligence in Cities and Regions, Mr. Erickson explores the application of crowdsourcing to cities and regions – what he called “geocentric crowdsourcing.”

In his opinion, rather than people as passive subjects of increasingly ‘smart’ systems developed by tech companies, the vision is that smarter cities can offer a variety of ways for humans to act as first class participants, contributing their abilities to sense, analyze and act. Mr. Erickson focuses his research on designing systems that enable groups of people to interact coherently and productively: originally focused on online systems, the scope of his work has expanded to include real world environments ranging from rooms to cities.

Responding to some questions from Crowdsourcing.org, Mr. Erickson explained the concept of Geocentric Crowdsourcing and his practical application to the cities.

According to the recent United Nations global report on human settlements entitled “Cities and Climate Change”, more than half of the world’s population now live in urban centers. Soon it will be two-thirds. This, coupled with an array of increasingly severe environmental pressures, suggests that cities and regions are an appropriate focus for R&D aimed at making them more livable and sustainable places.

In certain aspects this has already been happening, observed Mr. Erickson. Cities have been receiving increasing attention from purveyors of technology. However, in these visions, people are typically relegated to the periphery. They are often treated as passive subjects, as the recipients of the benefits of “smart technology.”

According to Mr. Erickson’s research, there are a variety of reasons for this. “One is that often these systems are being built by computer scientists and other technologists, and these professions tend to emphasize data over people”, he said. “Another is that often one can just go out and 'grab' the data by mining twitter, or the web, or what you have; if you want to engage people and make them part of the system, that requires a lot of work to make people aware, to convince them to participate, to motivate them, and (depending on how you motivate them) to deal with issues such as gaming the system”. Mr. Erickson completed: “Third, computer scientists and technologists aren't trained in designing systems that address human engagement of this sort. Fourth, it is only recently that it has been possible to do this, and so ordinary folks aren't necessarily used to thinking that they can participate in these ways”.

Mr.Erickson observes as well that over the last decade, an alternative, more social view of “smartness” has emerged. In this view, systems gain intelligence from both digital and human elements. People are seen as being actively engaged in systems as participants, and as a consequence can contribute their considerable knowledge and expertise to systems.

“Humans can contribute deep, qualitative knowledge; they can analyze fuzzy or incomplete data; and they can act in ways that digital systems often cannot”.

More generally, explains Mr. Erickson, when the basic task in the crowdsourcing system is simple, digital mechanisms can be used to integrate the work of the crowd. In contrast, when tasks are more complex, as in Wikipedia, digital integration mechanisms rarely suffice. More complex tasks are likely to be more intrinsically rewarding, and in addition, the social interaction required for integration and quality control is something that many find motivating in and of itself.

“A big issue here is that for such crowdsourcing to work, people have to come to believe that the system is fair -- that is, that it has not been hijacked by a special interest group”, he highlighted. “Part of the solution here is to be able to identify who is contributing -- that doesn't necessarily mean identify down to the name and address of a person, but if, say, we have a system that allows people to weigh in on ideas about city planning or how to allocate parts of the budget, you want to be able to show that you are actually getting a good cross section of people participating”. Mr. Erickson explained: “For instance, imagine that a system would be able to validate that a participant lived in a particular block, or neighborhood, or district, while not revealing exactly who they are. If I could see -- and feel confident that it was true -- that a broad cross section of folks was weighing in, that would increase the perception of legitimacy”.

A crucial question in this case is how to promote the interaction between the crowd and the responsible officials? According to Mr. Erickson, one approach is to involve cities in the development of the applications, so that the input of the crowd is compatible with their systems. Another is to promote data standards for describing and reporting various sorts of urban problems.

“The really tricky issue here is that the problem is not just reporting problems -- often there are shortages of funds, labor or other resources. How does one priorize, and do so in a way that is transparent and understandable to those who may not understand or be aware of the complex constraints of city planning under conditions of limited resources?” he asked.

Is his opinion, however, there is a number of reasons that crowdsourcing makes particular sense for urban contexts. First, that is where the people are – by definition, cities are dense concentrations of people. Second, inhabitants of a place develop a deep knowledge of it because they live, work and socialize there. Third, inhabitants have a practical interest in participating in systems that impact their daily life: someone who may never vote in an election may still complain vociferously about potholes in a local street.

“Finally”, he added, “inhabitants of a city or region often identify with it, or with networks of family, friends and communities associated with it; this pre-existing social structure is a valuable asset that non-localized systems like Wikipedia must develop from scratch”.

Mr. Erickson showed three examples of geocentric crowdsourcing.

Investigate Your MP

“Investigate Your MP” is a site run by The Guardian, a London newspaper, in the wake of a scandal about excessive expense submissions. The site invited citizens to analyze expense reports submitted by MPs, and to flag those that deserved closer scrutiny by the site’s administrators.

While this system does not tap into local knowledge – it only requires commonsense knowledge about what valid expenses are – it does tap into local motivation, leveraging public indignation about the purported misbehavior of their own representatives.

FixMyStreet

FixMyStreet is an application that allows citizens of the UK to report potholes, streetlight outages, and other street related problems on a publicly visible map. The problems are then brought to the attention of the appropriate council [governing body] responsible for fixing them. As individuals’ reports appear on the shared map, it creates a powerful aggregate representation of the state of the streets – areas with lots of problems become quite apparent.

Cyclopath

Cyclopath is a computational geo-wiki, a user-editable map that is used to compute routes, these computations making use of the edited map data. Cyclopath is intended for bicyclists, and enables them to find bicycle-friendly routes around the region; it relies on the cycling community to add data – road surface conditions, off-road paths, location of coffee shops – that is useful in determining a good bike route, but not found on conventional maps.

Cyclopath, which covers the 7-country metro region of Minneapolis –St Paul, has about 1500 registered users, who over the first 9 months of its use made edits to the map that resulted in routes that were approximately 8% (1K) shorter.

An interesting aspect of Cyclopath can be seen by contrasting it with Google Maps' recent introduction of bicycle routing, explained Mr. Erickson. Google Maps offers bike routing as a generic service; in contrast, Cyclopath is local – it relies on a place-based community to contribute local knowledge, and its existence and use is a point of pride for the local bicycling community.

“To me, this gives Cyclopath, a very different feel – and a very different set of social dynamics – from systems like Google Maps, without communities, or even Wikipedia, which has a community but one in which place is largely irrelevant”, he said.

Mr. Erickson believes that there are many other areas where geocentric crowdsourcing could be used in the same way.

“I think that place-based systems can work in domains which have some or all of these characteristics: there is an existing community or interest groups (or perhaps several) that care about the domain and there are concrete and small scale (individually-doable) actions that be taken that depend on local knowledge and local participation”.

He observes as well that the actions must be visible in the real and online world, so that participants can see both the results of their individual actions, and that others are acting as well. “There are tangible benefits to the individual or the community/group (or, ideally, both)”.