A typical automatic license plate recognition system uses infrared light to illuminate a plate even in the dark. A high-speed camera photographs the plate, which is analyzed and compared with a database. The system alerts officers to any matches. Whereas older scanning systems used a trigger mechanism, newer systems can run continuously while patrol officers are engaged in other tasks. These systems were enhanced in 2004 when the FBI's National Crime Information Center developed a process for providing agencies with a data extract of stolen vehicles, stolen plates, and wanted individuals that is updated daily. Uses for the system include locating stolen vehicles and identifying the car of convicted sexual predators who drive into areas prohibited to them, such as school zones. The technology still faces both technical and legal challenges. Technical challenges include differentiating between license plates from different States and dealing with obstructions. Legal challenges pertain to privacy concerns and possible misidentification. In the latter case, officers need to match not only the vehicle's license plate but also the identity of the driver and/or passengers before making a stop. In choosing a system, an agency must decide whether to select a portable system or one that requires permanent mounting. The choice will depend on where and how the system is to be used.