The US State Department “Tag Challenge” 2012 simulated a multi-city deadline manhunt. In London, Stockholm, Bratislava, New York City, and Washington DC each, a simulated jewel thief spent a staggered 12-hour period ambling about their city, appearance and attire unknown until the moment they set out. The goal was to find and take a picture of as many of them as possible within their 12 hours out and about.

The winning strategy used by our team, "CrowdScanner", to nab 3 of the 5 simulated targets, was preparation using a web platform developed to get members of the public to recruit individuals in their social networks into our network of on-the-ground spies, using a recruiter-rewarding incentive mechanism involving a combination of goal-proximity recruiter rewarding, as well as positional recruiter rewarding. Given the relatively tight geographic zones of search, it was also a test case for geographical directedness of social search.

We also worked on a project involving a less geo-directed search, analysing a similar recruiter-rewarded incentive mechanism used in the earlier DARPA Network / Red Balloon Challenge, using computer simulations.

Java/JSP + HTML/CSS + Apache Tomcat + MySQL

Masdar Institute of Science and Technology, 2012


Explore relevant MIT Media Lab project pages for the CrowdScanner / US State Department Tag Challenge here, and for the DARPA Network / Red Balloon Challenge study here.

Find media coverage of Team CrowdScanner's US State Department Tag Challenge win at Nextgov, The Economist, New Scientist, Popular Science, Federal News Network, The National, ZDNet, and Vision; of the extended crowdsourced manhunt study at MIT Technology Review, Scientific American, New Scientist, and Nature; and of the DARPA Network / Red Balloon Challenge simulation study at NBC News, ABC, Arabian Gazette, and Gulf Today.

The winning incentive mechanism

The winning scoreboard

Coverage in Nautilus magazine