Walt Disney Imagineering 1980
1980 Big Thunder Mountain Railroad Walt Disney World
Well Big Thunder at Disneyland just opened and it was time to start on Big Thunder for Walt Disney World. Harry and I were having lunch in the WED cafeteria one day and Dick Nunis, Chairman of Walt Disney Attractions sat down and starting talking to us. Now Dick had a reputation of being a bigger than life personality. We heard stories that probably were not true that we would just fire people on the spot. The saying "no news is good news" became "No Nunis is good Nunis". Well, to us he was the nicest guy. He explained that "you have me to thank" for Big Thunder opening next September. I decided not to open both at the same time but to stagger them to give you all more time and to learn from the first and apply it to the second. He talked a while longer and got up and left. That was a wonderful moment with someone many levels above us and this would not be the last time that would happen.
I will pause here to say that all my previous companies were Engineering companies and you addressed all people above you as mister (sorry they were all men). So Mr. Nunis or our VP of Engineering, Mr. Zovich. Well at Disney you address everyone by there first name and they all wear their nametags. I think we were too chicken to address Mr. Nunis that way that day but as time went on, I got used to addressing everyone by their first name, even most recently addressing Mr. Robert Iger, CEO as Bob.
We convinced management to let the software people, us, rewrite the software for Florida's Big Thunder. We would use my compiler and approach the problem from a new point of view. We started from scratch with the software but had the basic algorithms from Disneyland. The PLC had no subroutines so when you had say 10 sensor zones with the same software you have to enter it 10 times with the different sensor names (actually a number there were no names). This was prone to typos. With the macro-compiler we wrote one macro and called it 10 times with each sensor names (we now had names!). It created 10 identical copies of the code with only the sensor changed as appropriate. It actually only took us only 2-3 months to write and test the software.
Big Thunder was the first and only ride system with a built-in simulator that simulated all hardware inputs. We sat in front of the actual Ride Control system cabinets and ran our software testing all five trains running around the (simulated) track. You could cause failures of any piece of equipment and the test system response.
So we finished everything by February and were not scheduled to go to Florida until June. The first part of any installation is to test all the individual pieces of equipment as they are installed. So, for example, test that every sensor when flagged is seen by the computer. Manually tell the computer to open a brake, and so one. Only after everything is installed and the mechanical engineers test that the train physically makes it around the track does real Ride Control testing start.
So our real testing started in August with a mid-October planned opening. It had been seven months since we wrote the software and we had forgot how we had program some of the operations. The Ride Operators were there with us testing and once in a while we would say to each other, hey what is the sequence of buttons we press for this operation? Well you could see the look on the Operators faces like do these guys know what they are doing? Well a few days into testing it all came back to us.
We were asked if we could open the attraction on September 23, some 4 weeks early which we did. This was a big deal to us because the last three rollercoasters had all open later than originally plan for various reasons.
I'll pause again to say that Disney will open an attraction before the advertised opening, this is call the soft open. If you happen to be in the park that day you may get a sneak peek of the new attraction. This allows Imagineers, Operations, and Maintenance, a dress rehearsal of sorts to iron out all the kinks.
The grand (advertised) opening was in October and this attraction, only the second roller coaster ride at Walt Disney World became very popular with the guests.
Side note: This software used a new philosophy I took from "Logicon: A Flash of Enlightenment". Most safety software looks for error conditions to stay safe. The creator lists all the thing that can go wrong. The problem is what happens if you forgot one? This different approach is to write software that says there is one and only one safe way thru the attraction, everything else is an error. In reality the final software for both methods are similar but different enough to make a difference. Years later we hired a famous company to look over a proposal to update the software for an older ride system. We explained this method and they used that in there report to Disney as to why it was a good idea to update this software. Yes we paid $100K to educate a consultant on our ways of writing software. During this time this method became known as the "one safe way down the mountain" method.
Back to Disney Years