CognitecArticle2Image1 CognitecArticle2Image2

Australian Custom’s SmartGate Kiosk which uses Cognitec’s FaceVACS face recognition software has been successfully operating at Sydney Airport since November 2002.

More than 4,400 aircrew are using the kiosk as an automated passport control system on arrival from international flights. Since February 2005 this has been extended to include enrolled Qantas Platinum Frequent Flyers returning from overseas.

Customs SmartGate is an automated border control system that undertakes the face-to-passport check usually undertaken by a Customs officer. The system uses face recognition technology to compare a live image taken at the Customs control point against one or more stored images of that same person.

In a world first, SmartGate has been installed at the Customs control point for Qantas aircrew at Sydney International Airport. In less than 10 seconds SmartGate performs all necessary image verification and customs/immigration checks and allows enrolled aircrew to pass through the Customs control point quickly, while maintaining a high level of security.

All users must have previously enrolled in the system. During enrolment Customs validates the travellers passport, performs the face-to-passport identity checks, scans biodata from the passport and collects five images of the person. The process takes less than one minute.

The only images used during the SmartGate transaction are the enrolment images and the live images of that same person taken at the kiosk. These images will not be compared to images in any other database.

SmartGate enrolment is entirely voluntary. If Qantas crew do not wish to use SmartGate a Customs officer can manually process them at the crew channel.

A survey among Qantas aircrew users of the system showed that 98 per cent said they preferred the machine to a check by a customs officer.

The independent evaluation of the SmartGate project, conducted by Dr Jim Wayman and Dr Tony Mansfield, advises that there are no technical reasons why it could not be expanded to other groups and locations.

“SmartGate’s performance is remarkably good for an operational facial recognition system,” Dr Wayman said. “We know of no other face-recognition system with documented performance at this level elsewhere in the world.”

Drs Wayman and Mansfield reviewed and reported on three reports that form the basis of the study. The Defense Science and Technology Organisation (DSTO) conducted a technical trial of the system, ACTSAFE Australia conducted an ergonomic assessment of the system’s features and ACNielsen conducted a user’s survey. They also conducted independent observations and reviewed SmartGate data logs, internal memos, press coverage and other reports on the project.

“In-house performance data, collected during the SmartGate trial, always indicated that the system was performing at better then expected rates. It is extremely satisfying that the work by Doctors Wayman and Mansfield, DSTO, ACTSAFE Australia and ACNielsen confirms our findings,” said Gail Batman Customs Border Intelligence and Passengers National Director.

In May 2003, ICAO (International Civil Aviation Organisation) adopted a globally harmonised blue print for the integration of biometric identifiers into machine-readable travel documents (MRTDs). Face recognition was selected by ICAO as the globally interoperable biometric for machine-assessed identity confirmation. This is consistent with current practice and the widespread use of “photo ID” to verify identity during the travel process.

Alfredo Herrera, Managing Director of Cognitec Systems, said: “Our team at Cognitec is very satisfied with the results of this study which confirms the outstanding performance of our technology. For us, SmartGate has been a model for future border control systems compliant to the recent recommendations by ICAO. Various countries are planning for similar projects and we feel well prepared to reply to their requirements.”

See full article at–cognitecs-border-control-software-in-successful-pilot-at-sydney-airport.html

Leave a Reply