Five steps for identifying the right biometric ID system for your company

Biometrics offers unprecedented security that can be deployed to safeguard your data and physical premises, but choosing the best system to invest in can be a challenging task

According to an article by CIO Insight, the cost of biometrics security when safeguarding your physical premises, for instance, averages between $2,500 and $10,000 per door.

With that in mind, it’s important to know the level of investment you need to make when looking at biometric security systems, and the first step in that process is understanding what those systems actually are. 

There are several basic types of biometrics systems available, so picking the right one requires knowledge of the basics. You don’t need to implement expensive, airport-level security measures to replace the keycard system that lets you through the turnstile of your building, for instance. 

Let’s take a look at a brief overview of each system and what they do.

ABIS – Automated Biometric Identification System

First on the list is ABIS, which stands for Automatic Biometric Information System. Historically, this type of database has been used by the police, so if you need to keep track of thousands of fingerprints or faces, this might work for you.

For instance, Mexico has a regulation that requires banks to use biometrics to verify the identities of their clients as an anti-money laundering measure.

ABIS is expensive, but if it’s something you need (when the law requires you to check the identity of clients securely), it can be good for the risk management. In the United States, companies that need to adhere to anti-money laundering statutes can use an ABIS system to minimize liability. 

One cool feature that ABIS incorporates is how it uses biometrics to compare a selfie with your ID and then back check your ID against a government database to make sure it isn’t stolen.

In the Philippines, you can get a fake ID for around $50, and fake IDs are endemic, making identity verification unreliable. One of our clients in the Philippines works by giving out small consumer loans given out after collecting fingerprints. The bank runs fingerprints to see if they can give out a loan to a customer based on their credit history. If they show up in the system as a different person than the name on their ID, the bank is able to eliminate fraud. In fact, biometrics helped this client cut down on fraud by 90%. 

On the other end of things, banks can qualify people for larger mortgages if they know their information is very safe and tied to a biometrics database. This makes consumer loans cheaper.

Remote Onboarding 

The next type of biometrics system is remote onboarding. This can be used by any business that sells a service that requires customers to fill out many forms. With digital onboarding, you can use your smartphone to take a photo of your face and an ID to sign up for a bank account, log in to certain government websites, and eventually do things like pay utility bills.

These systems use OCR – optical character recognition – which means you take a picture of your ID, and the system reads all the text on the ID. It then takes that data and form-fills for you when signing up for an online service. 

Largest Slovak bank uses this even in the branches to help you open an account in two minutes rather than half an hour. It’s so fast that now you come to the bank branch and you sign up using an app on a tablet. “You can’t make the process simpler,” one bank representative said.

Facial Recognition Access Control

This is a good one if you’re looking for physical security. When you enter a building, Facial Recognition, Access Control software, implemented into your CCTV system, assesses your face, gait or physical presence and compares it to a database. Based on your physically distinguishable features like the way you walk, paired with the way you look, the system will automatically open the turnstile for you or call an elevator to the floor you work on. This can be used to only allow certain people access to the floors they are permitted to access. 

Facial Recognition Access Control measures can also be paired with your in-office attendance system, so you know when people came into work, when they left from work, and when they took their lunch.

In some countries, this is particularly important. In Dubai companies have problems with lots of migrant workers who would share the same work permit. Facial recognition makes sure that only the individual permitted to walk in can enter the building. 

This type of system could be deployed in airports for biometric airport tickets. It’ll be connected to your biometric IDs so it will allow you a customized journey through the airport. All you would have to do is simply look into the biometric kiosk and it tells you your gate number and where you need to go.  At places where heightened security systems are needed, this brings immense value in shorter waiting times and reducing bottlenecks.

Visitorship Analysis

This type of system is excellent for large-scale data collection about visitors in venues. Visitorship analysis can analyze data from CCTV on an anonymized basis to estimate age and gender to figure out the number of people in a large venue (like a mall) and their demographics. After this data is collected, the system automatically throws away the faces when the people it is tracking leave the venue. 

This allows event coordinators or large venue owners to track customers’ faces for the duration they’re in the mall or at a festival. This can be used to gather age/gender cohorts: for “Today there were 5000 males and 7000 females ages…” for instance, and property owners can tailor their offerings to that particular audience.

So, let’s say the owner of a shopping mall needs to attract more of a certain group, and they’re reviewing retail real estate applications; they might give the lease to a store that can attract that desired demographic.

Finally, every respectable biometric algorithm is categorized by the National Institute of Standards and Technology (NIST). Every company that is important in the field of biometrics sends their algorithm to NIST, and every company that means anything in biometrics is there, ranked, by their accuracy. For example, NIST’s Facial Recognition Vendor Test (FRVT) can help you pick out the best facial recognition vendor.

When making your final decision, it’s important to ask two questions:

1. How do you deal with personal data? For instance: Is your data stored all on one server? Or compartmentalized so IDs and faces are kept separately? Also, what are the compliance standards (ISO or otherwise) you adhere to?

2. Does the algorithm you use identify people fairly? For instance, does it contain racial bias? Can a potentially biased algorithm affect the reputation of my company?

The best biometrics companies offer secure storage of your data under your full control and can identify people of all racial backgrounds even when wearing a Covid-style facemask. Don’t let vendors fool you into thinking this can’t be done. Again, NIST has made reliable measurements of both bias in facial algorithms and their ability to recognize people in masks.

Whether you are looking for a risk management solution or an attendance system, biometrics can make them more reliable, more comfortable and more inclusive. Many companies now realize that biometrics, if used properly and in accordance with personal data protection, can be a game-changer, and it doesn’t have to be out of reach for everyone. 


About the Author

Jan Lunter is CEO & CTO of Innovatrics. Graduated at the Télécom ParisTech University in France. Co-founder and CEO of Innovatrics, which has been developing and providing fingerprint recognition solutions since 2004. Jan is an author of the algorithm for fingerprint analysis and recognition, which regularly ranks among the top in prestigious comparison tests (NIST PFT II, NIST Minex). In recent years he is also dealing with image processing and the use of neural networks for face recognition.