Big Data Fights Crime: The FBI’s Next Generation Identification

The Federal Bureau of Investigation has been in the Big Data business since 1999 when it launched the Integrated Automated Fingerprint Identification System (IAFIS), the world’s largest biometric database on individuals. IAFIS contains over 55 million sets of fingerprints and is used by law enforcement agencies nationwide for criminal investigations and civil background checks. To keep pace of the evolution of biometrics as well as the ever-growing demand for information, the FBI is now incrementally upgrading to a system it calls Next Generation Identification (NGI). Through cutting-edge Big Data solutions, NGI promises to improve service dramatically while at the same time increasing functionality.

As the FBI boasts on its website, NGI will be “Bigger – Better – Faster” than the current system. It will incorporate the Department of Homeland Security‘s IDEN database used for checking whether those booked by local law enforcement agencies are in the United States legally, adding millions of records. The biometric information NGI will contain will also expand dramatically to include palm prints, iris scans, and mug shots that can be searched using facial recognition technology. The service it provides will also be improved through capabilities such as the Rap Back Service, which allows users to get updates on new criminal violations after processing, and an automated quality check, bringing the proportion of transactions requiring manual review from 98% down to about 15%.

But what’s most remarkable about this new technology is that it provides this additional functionality with greatly improved speed. Currently, IAFIS takes 2 hours to return results for fingerprints from the 2 million person Repository for Individuals of Special Concern in a criminal case and 24 hours for a civil investigation. NGI will be able to process high priority criminal requests in 10 minutes and high priority civil queries in 15, and takes only 30 minutes and 2 hours for routine requests. Low priority requests take a day and non-urgent requests could take longer as time is not a priority.

The software that allows the NGI to perform these millions of calculations quickly is MorphoTrak, which combined innovative biometric algorithms with some of the current best practices in Big Data solutions. The algorithms were ranked first in the National Institute of Standards and Technology‘s 2001 Evaluation of Latent Fingerprint Technologies for accuracy. MorphoTrak’s software has service-oriented and highly distributed architecture built on commercial-off-the-shelf hardware and complies with international and industry standards so it’s highly scalable and easy to upgrade and modify. MorphoTrak also supports data interoperability and uses Oracle Database 11g to support clustering and private clouds. Together, this allows for both better and faster searches, helping rapidly identify suspects and speed up check backgrounds to aid in the FBI’s law enforcement mission.