From slippers to biometrics

Image from Pixabay

Some nursing homes use facial recognition to keep elderly people with dementia inside, the Dutch tv news reported a few months ago. Because I am always on when it comes to possible topics for this blog, I made of note. And now I finally get around to explaining why that report caught my attention.

Facial recognition is a form of biometrics, just like a fingerprint scan or voice recognition. Biometrics means something like 'measuring biological characteristics'. The technology is based on the fact that every person has a number of unique characteristics. Based on these, you can identify someone. And to reassure you: biometrics doesn’t store your complete fingerprint or a photo of your face. Instead, a number of specific characteristics are recorded, such as the distance between your eyes and other proportions. When checking your access rights, a camera or scanner is used to check whether these characteristics are in its database. That is why the fingerprint scan on your phone suddenly works less well if you have been doing a lot of DIY: your finger is too rough to match.

So we use biometrics to gain access to something. Not to be denied access. But that is exactly what those nursing homes do. The front door is always open, but if the camera sees someone approaching who is not allowed outside because it is not safe for them, the door is locked. The nursing homes love it: "Otherwise we have to keep the doors closed for all residents. Now we turn that around: the doors are open."

And what if a smart resident sticks on a fake moustache, I wonder. Or puts on sunglasses. There is a good chance that he will not be recognized and will happily walk outside. Now I don't know if smart and demented can go together, but yes, I am obliged to my position to assume that things can go wrong. Edward Murphy is my role model (you know, the one with that law: everything that can go wrong, will go wrong).

What we see there is biometrics turned upside down. Why is biometrics not applied in the usual way? Everyone who is allowed to go outside is in the system. If he or she is recognized, the door swings open. If someone comes shuffling along who is not allowed to go outside and therefore is not in the system, the door stays closed. You have to be very clever to fool the system.

Before those nursing homes switched to biometrics, they used wristbands or sensors in their clients' slippers. Even then, they worked with open doors, which were locked only for some. But of course, you could easily work around that: take off your slippers and voila, you were outside. And a bit of fiddling with the wristband also turned out to work. Incidentally, the switch to biometrics has a double face: on the one hand, a band that is visible to everyone has a stigmatizing effect, on the other hand, the barely visible biometrics makes it difficult to enter an official protest – a right that also dementia patients have.

A nursing home is not a prison. Only residents who, due to their condition, are not safe to go outside alone, are kept inside – with the permission of themselves or their legal representative. Visitors are welcome and must be able to walk in and out freely. Open doors give a relaxed feeling, and thus contribute to a dignified existence. From that perspective, I understand the reverse approach, and I can imagine that there will not be that many clients who know how to hack the system. For most other applications, however, I like to stick to biometrics as they are intended.

 

And in the big bad world…

• You will soon be looking at advertisements featuring yourself on the lock screen of your Samsung phone.
• Twitter/X's new secure DMs aren't really safe.
• Microsoft hasn't turned off the International Criminal Court's email after all, Microsoft says.
• Cybercriminals sometimes attack their own kind.
• you should beware of fake versions of Zoom (and similar software).
• AI companies are moving too fast, at the expense of our safety.
• Facebook and Instagram spied on you on your Android phone. [DUTCH]
• You can read much more about how you were spied on here.
• This video warns about deepfakes.