San Diego Fireworks Display
Table of Contents
San Diego Fireworks Display
Summary
On New Year's Eve 2012, due to a programming error, a coordinated 17.5-minute firework display was instead set off over a period of 25 seconds. This caused shock as multiple large explosions were heard simultaneously across the city. The company involved, Garden State Fireworks , has little to say about it on its official site, so I am trying to track down any information about what exactly went wrong.
| Date | 2012-12-31 23:59 West Coast |
|---|---|
| Cost | $400000 |
| Affected | 500000 people |
| Deaths | 0 |
| Injuries | 0 |
Analysis
While this was obviously tested, the real run did not go as anyone expected. An organiser blamed it on a "technical glitch". We will find out what that actually means. It seems likely this is because the real explosives were set off by the program initialising. An "unintentional additional procedural step"?
"A set of instructions that were given that we didn’t necessarily create, that was created by the system." - A. Santore
The problem was a millisecond instruction that you couldn’t see with the naked eye - A Santore
A likely story:
The program was scripted by our company specifically for San Diego to be choreographed to the music. The computer reads it and sends signals when you want something to fire. - A Santore
We sent the signal to start the program running. At four minutes, it came to a first test to light an igniter at each location so we see a little match pop to see everything's fine. But instead everything opened up. All the masters fired. The pyro worked 100 percent. Unfortunately it all worked at the same time. - Santore
We accept the responsibility. We don't accept failure and we'll get to the bottom of it. We certainly feel bad for all the people who went out there to see it. It would have been the most spectactular show San Diego ever had.
- A Santore
Looking at all these quotes, plus some video material, it looks like there was a low-voltage signal to confirm connectivity, and a test signal which had enough oomph to actually set off the charges: could it have been that the targets for these two were switched, resulting in a positive connectivity check, and a catastrophic failure to do a partial ignition check?
It's not clear from any sources so far what software they would have used, so it's hard to check how plausible that kind of mistake would be.