AI and Biometrics for a Smarter City

Biometric technology along with AI are paving the way in identity verification procedures and mitigating security risks.

Overview

With advances in biometrics technologies and improvements in IT infrastructure, there is a growing acceptance of biometric recognition technologies in our daily lives, and this acceptance will grow further with time.

Ian Mccarty

Chief Technology Officer / fairsystems
Biometric technologies are presenting new opportunities for organizations, both public and private, to create new innovative identity and access management solutions. With the growth of globalization, increased mobility, and the societal shift toward online activities and home-working, the pressure is on for biometrics technologies to provide more robust forms of authentication to better facilitate business and citizen transactions, protect organizational data, and facilitate travel and trade.

The Client

Our client, a policing and border control agency, engaged with us to deploy an AI-enabled facial recognition solution to screen and regulate passenger flow through its border crossing checkpoints. A legacy border control solution was used to enable physical access control primarily but our client was no longer satisfied with the accuracy and speed of the existing solution. In addition, our client wanted to identify persons of interest and high-risk travelers.
Roadmap
1
Assess
The initial phase of this project included an in-depth assessment of the existing systems and technologies being actively used by our client. Next, came a detailed analysis of the capabilities and limitations of these systems and their compatibility with the future design of the AI-powered face recognition solution. The result was a detailed business case covering the future system design and recommendations.
2
Deliver
The system was deployed in Automated Border Control (ABC) areas supporting border guards to facilitate smooth and non-intrusive rapid crossing by travellers based on deployment of the next generation biometric identification detection methods. The ability for the system to efficiently process low-risk travellers, combined with increased levels of accuracy, security and privacy standards and enabling border guards to concentrate resource on higher-risk travellers, was a key target of the project. To achieve these goals, a multi-biometric enrollment and verification system was envisaged, taking into account current and next-generation e-Passport chips, mobile equipment and person identification ‘on the move’ using facial recognition technology.
3
Continue
A detailed analysis was undertaken to optimize the currently deployed biometric modalities, application of emerging biometrics, multi-modal biometrics, and counter-spoofing, for border control scenarios. An integral part of the project was the collection and dissemination of new border-realistic biometric datasets and systematic evaluation of the developed biometric methods including vulnerability and privacy assessment. The project was strongly user-driven and demonstration of the developed biometric system was be conducted at two different border-crossing sites.
Solution Details

@fairsystems, we know that challenges are merely opportunities in disguise.

Biometric data capture with quality assurance
Feature: The solution detects passengers' biometric information, including gender and age, and stores the corresponding biometric profile only if the quality of the detection meets a predefined threshold.
Biometric traveler verification (1:1) and identification (1:N)
Feature: The solution matches and compares passengers' biometrics against travel documents, existing profiles, passport photos, and national and international monitoring lists.
Automated travel document identification and verification
Feature: The solution detects potentially fake travel documents by detecting the presence of anomalies and modifications. In addition, it compares passport information against a centrally administered document repository.
Integration with national database
Feature: The solution integrates with the national identity and visa information system for visa status checks, verification of visa documents, and biometric identity confirmation.
Automated comparison of identity
Feature: The system uses ID document verification and facial comparison together to verify the authenticity of the ID document and the presence of the genuine owner of the passport in real-time. During this process, a trusted source image of the passport holder is compared with the real-time photo of the person trying to get through the gate.
Visa entry/exit and transit monitoring
Feature: The solution provides an advanced transit alarm logic powered by AI and rule-based analytics to determine visa overstayers and potential security risks.
Compatibility with legacy systems
Feature: The implemented solution was specifically selected to be compatible with existing products, technologies, and systems, making the implementation project non-disruptive and reducing adoption costs.
Profile-based identification and storage
Feature: All events with the same individual are automatically combined into a single entity — person, which significantly simplifies the search for the necessary information collected at different times and under different conditions.

Result:

Today, our client’s border management system is deployed at 25 border points (including 2 international airports and 3 land border points), with 118 primary inspection stations and 31 secondary inspection stations. The system has the ability to detect faces in a video stream in real-time, accurately count people, check faces against monitoring lists and Interpol STLD, notify the user when matches are detected, and provides informative statistical reports.
0

Checkin time
decrease

0

Passport control
time decrease

0

Security checkpoints
time decrease

0

Boarding process
time decrease

0

Monitoring list
accuracy increase

Testimonials

How our services bring about success