3 ATM Questions

Once upon a time a customer of my employer misconfigured his ATM. It was programmed, somehow, to think that the $20 bills in its casette were $2 bills.
Take a guess what happenned when you asked for $60?
It obligingly handed you 30 of what it believed were $2 bills.

I actually opened this thread hoping for a brisk discussion on the merits of packet switching (Asynchronous Transfer Mode).

OK, I’ll just crawl back under my rock…:smack:

I see a problem with the tally up method of checking if someone has been short changed.

The above stories mention both, people who receive too little money, and people who receive too much. It will generally only be the people who receive too little who report it. If you have someone overpaid $20.00 and someone else underpaid $20.00 on the same day, the ATM totals will look good at the end of the day and wont prove that person 2 got underpaid.

I used to work for a software company that would process the transactions on the bank side.

Without all the gory details, there are two parts to a transaction. There is a request for funds, and a completion on that request. The first part puts a hold on your account, the second actually debits the money. (part 1: someone says I want $20, the ATM finds out how much money you have, and puts a hold on your account for that amount. Part 2: the person gets the money and the ATM sends the bank notification the money was received). The two are matched up by various things, such as time of the transaction, amount and account number. Matching them is crucial.

Well, someone (more than one someone) figured out that doing back to back transactions would mean the second one was not debited. Someone, if my memory serves, did it to the tune of over $40,000. More than one someone, too.

Here is a link to verify the story outline. I can’t find verification of the $40,000, though. However, I was very much a part of the company that dealt with cleaning this up. In other words, this isn’t a friend of a friend situation.

No, it wasn’t my bug.