The number of people scanned through the gate has to match the number the cabin crew count in the seats, or the aircraft does not move. On gate 6 they disagree by one, about once a fortnight, and it is always the scanner that is higher.
I am a dispatcher. Nine years on the same pier, and before that four years on the desk.
The count works like this, and it is deliberately two counts taken by different people who cannot see each other's answer.
At the gate, every boarding pass is scanned. The reader keeps a log with a time stamp and the sequence number off the pass, and that gives me a number I can say out loud.
On board, the cabin crew count heads. Two of them do it independently, front to back and back to front, and they count infants separately because an infant on a lap is a different line on the loadsheet.
Then the senior brings me a figure, and I have mine, and they match, and the doors close.
When they do not match, nothing happens until they do. You recount. If it still disagrees you do a boarding pass check, which means walking the aisle and looking at every pass against every seat, with a hundred and seventy people watching you. If that does not resolve it you offload bags and start again, and that is a flight gone for the day.
The honest causes are dull and I have seen all of them. A pass scanned twice because the reader double beeped. A staff traveller on a jump seat. A passenger who got up to use the toilet during the count and got counted in two places, or in none.
On gate 6 the scanner comes out one higher than the crew.
About once a fortnight. Always one, never two. Always the scanner high, never the crew.
I have looked at the log on twenty three of them now.
The extra scan is a real record.
That is the part I would want somebody to argue with me about, because everything else follows from it.
There is a line in the reader log, with a time, and it carries a boarding pass sequence number. A sequence number is issued by the check in system when the pass is printed, and they run in order for that flight.
The number on the extra line is one above the last number the system issued for that flight.
So it is a pass for that aircraft, next in the sequence, and the system that issues them has no record of printing it.
What has been done.
The reader replaced, twice, and the second time with a different model on a different software version.
The gate's network drop and its session handling looked at by the airline's IT, because a duplicate submission is exactly the sort of thing a flaky connection does, and they found nothing and said so in writing.
A manual tally run alongside the scanner. This was mine, and I have done it eleven times and asked two colleagues to do it on their own shifts.
Twice, the manual tally has also come out one higher than the crew.
That means a person with a clicker, standing at the door looking at the bridge, counted somebody through who was not in a seat two minutes later.
I have asked both of them. Neither remembers anybody unusual. One of them, who I have worked with for six years and who is not fanciful, said she thought she had clicked twice by mistake and then said, when I pushed her, that she did not think that.
How it ends, every time.
We recount. The crew count again and get their number again. We do the pass check, and every seat matches a pass, and every pass matches a name on the list.
The captain accepts the crew figure, because the crew figure is the people who are actually on his aircraft, and that is the right call and I would make it too.
It goes in the log as a boarding count discrepancy, resolved, cause gate reader. Twenty eight minutes, averaged across the twenty three.
The airline gets a delay code that puts it on us and we wear it. Four hundred pounds a pop or thereabouts in compensation, which over nine years is a number I have never been asked about.
The one thing that is not in any of the paperwork.
The time stamp.
For the first few years the extra scan sat anywhere in the boarding, middle mostly, the way a double read does.
I went back through them in order in the spring because the log keeps everything.
The last nine are all the final scan of the flight. Not near the end. The last one, after the genuine last passenger, after the final call, after I have said the words into the handset and looked at an empty gate room and turned round.
It is between forty and ninety seconds after the one before it, and the one before it is always me walking down the bridge to the aircraft door, and I am the only person on that bridge.
So I have started doing something small. I scan my own pass-through card at the reader when I walk down now, on purpose, so there is a line with my name on it between the passengers and whatever the last line is going to be.
On the two occasions since, the extra line has come after mine.