U-2 is a spy plane developed in the 1950s by Lockheed. They designed it for spying Soviets in the cold war.
What is unique about the plane, is its light structure and the high maximum altitude it can reach. It can climb to over 100,000 feet (~30,5 kilometers).
An altitude that high is not typical for our more modern planes. Los Angeles airport learned it the hard way. The flight control system was not designed to handle altitudes that the U-2 reaches and the system malfunctioned.
A simple software bug caused the system to interpret extreme altitudes as really low. Just an overflow in handling long numbers. It forced the flight control into a chain reaction of readjusting the flight routes of every single plane in southern California.
The system could not handle the load, and all flight control in southern California was left paralyzed for an hour.
Los Angeles airport only had to cancel 27 flights, reroute 27 to different airports and plan delays to 212 more flights. And other airports in the area had to do the same.
I always get curious when reading about this kind of incidents. My first impulse is to try and calculate the expenses for such events.
A little bit of Googling finds out that a flight from L.A. to New York could cost an airline $56,400. The expenses are mostly there whether they do or cancel that journey. And then there is the loss of revenue that would add up.
27 such cancellations would already cost the airlines $1,500,000. I wonder how many similar scenarios related to this one incident we could come up with if we tried?
But there is one more calculation to do before we head off. How much software testing could we have bought with that money?
An expensive test consultant would have worked for 1350 days straight with a fee of $150 an hour.
It’s curious to consider just how many bugs we could have caught with that kind of an investment.