Biometrics on ATMs

Biometric systems have been in use since the 19th century and have gained so much popularity over the years. Biometric systems utilize a human’s physiological, behavioral, and psychological traits to perform authentication and authorization.

Different kinds of biometric systems exist such as:

  • Iris detection & recognition systems
  • facial detection & recognition systems
  • voice detection & recognition systems
  • And fingerprint recognition & detection systems.

These systems have basic deployment phases which include enrollment of traits, matching of traits, and maintenance & support.

A generic biometric system has a sensor which collects the biometric traits, the traits are then pre-processed and the biometric features are extracted to generate the biometric templates. A biometric template (also called template) is a digital reference of distinct characteristics that have been extracted from a biometric sample. The templates are then stored in a database or any other form of storage, and they are always referred to during the matching phase. The figure below depicts a generic biometric system.

Biometric systems utilize the evaluation metrics for accuracy check, convenience, and performance. False Acceptance Rate (FAR), a measure of system’s accuracy is used to compute the percentage of the number of times in which the system granted access to incorrect individuals. False Rejection Rate (FRR) is used to compute the percentage of the number of times in which the system rejected the correct individual. Failure to enroll Rate (FTER) is utilized to compute the percentage of the number of times in which the biometric traits were not enrolled in the system(s), along with the Retrial Rate (RR). The performance of the system is then measured using the performance curves such as the Receiver Operating characteristic (ROC) and Cumulative Math Characteristic (CMC).

Biometric systems have numerous advantages such as producing great performance, cost efficiency, and convenient authentication, however they have numerous disadvantages which may prove to be severe. The first disadvantage is that once the biometric features are breached, they cannot be altered as they can in password authentication systems. The noise levels on the scanner may also introduce invalid features.

Livens detection of the features may also prove to be deadly if it is done incorrectly on the system. Liveness detection can be simply defined as the ability of a biometric system to distinguish the difference between the spoof and a real person. Biometric systems have additionally been criticized by users because of religious beliefs and their intrusiveness.

Despite the disadvantages that come with using biometric systems. Biometric systems are widely used in various platforms such as banking, law surveillance, time attendance, and many more. It is noted that First National Bank (FNB) is in the process of using biometric Fingerprinting on ATMs to secure the funds of its customers.

ATMs (Automatic Teller Machines) have played a significant role in transforming the lives of people in South Africa and the world as a whole. A substantial number of people in South Africa utilize ATMs for daily banking and they have become extremely dependent on ATMs. Imposters also depend on ATMs to perform various fraudulent activities such as bank card cloning, card trapping, false keyboards for capturing the PIN numbers of ATM users, and many more.

ATM security is the first priority of banks in South Africa and the entire world. The banks are forever seeking better ways to strengthen security on the ATMs. Although using biometrics on ATMs is costly, Biometrics could improve safety on the ATMs in an enormous way. Each and every system has its pros and cons, and I believe that biometrics is the way to the future.

Written by: Jonathan Baloyi

Image: Figure 1: A generic biometric system (Du, Eliza Yingz, 2013).