I don't have a full answer but what I observed is that there is always someone that is nearby and "in control".

The base I think works like this:

- you scan the passport

- you look at the camera

- your passport details & photo are sent to the control room with your live feed

- a guy looks at everything those details plus your immigration history and other info

- same guy decides to let you in or send you to a desk

This was particularly obvious in the old Gibraltar border where you could easily be the only person going through and so the agent had to do the routine just to let you pass.

Yeah this is pretty much my assumption. I would imagine that the person doing the judging may not be on site as it's probably cheaper and easier to have a bigger workforce in one office, flex the people in shifts, load balance across all ports with the gates, etc.

I do wonder how much is automated. For example, is the gate fully automated in calculating a risk score and then referring you to a border agent if above a threshold? Is a person looking at the details in realtime to decide that, or are they just doing facial recognition, or are they only involved when the gates fail.

Those stupid gates have never worked for me, and I've always suspected that they are actually just connected to a call-centre type operation. Given how fast my phone can recognise my face (in 3d no less), the e-gates are a total joke.

That's so weird, I use them regularly and have never had a problem in the UK or other countries with them.

I wonder if you're failing due to not recognising your face, or (no offence intended) that you have too high a risk profile in some way due to your passport info, and they decide they want to see you in person.

I have always assumed face recognition because my passport picture has no glasses and the gates never work when I wear my glasses. A human would recognise me, I think.

Interesting. I'm the opposite, my passport photo doesn't have glasses, but I've never taken them off for the gates and I don't think they tell you to, and I've never had a gate fail.

They do tell you to take them off. But they might be targeting shades more than regular glasses.

Oh! How weird. Maybe I'm getting the UK gates mixed up with Australia where I'm often travelling to. I do remember looking for a sign and being surprised that I didn't need to take them off recently, but it might have been outbound.

I don't know anything about this particular case, but about these systems in general. They can work without any operator or network connection. They verify that you have a valid passport and that the taken image (face) matches the one stored in your passport. But for modern passports they use Extended Access Control which requires up to date terminal certificates to access the data (you have to update them in the range of days) und you can give these systems revocation lists and lists of unwanted persons. If any of this is not updated, they stop working.

> But for modern passports they use Extended Access Control which requires up to date terminal certificates to access the data (you have to update them in the range of days

Passports don't know the current time and thus can't tell whether the presented certificate is within its validity range (as in a malicious attacker could feed an expired certificate as well as a fake "current time" value to make it appear valid), so why are those certificates short-lived?

Whenever you present a certificate to the passport, its current time is updated by the "valid from" value if it's newer then the current time.

It's not perfect but if you started you trip in another country with such a system and where a more recent certificate was used, your passport will deny access.