ELECTRONICS
Attraction hardware fails on repetition, not on load. Maintenance-aware design, and the parts that should be replaceable in the dark.
Hardware built for an installation is usually specified against the worst moment it will face. Peak current, peak temperature, peak load. That is the right instinct for a device that runs occasionally and the wrong one for a device that runs all day, because in an attraction the peak is not what kills it. Repetition is.
A connector rated for hundreds of insertions is fine until someone has to reach it every service interval. A fan is fine until it has run continuously for a year in a room with theatrical haze in it. A relay's specification is honest about its cycle count and nobody multiplies that figure by the number of shows per day until the day it stops.
The environment does its own damage, quietly and on a schedule nobody wrote down. Theatrical haze settles on optics and on fans. Fog fluid is mildly corrosive over a year. Temperature in a show enclosure runs higher than the room it sits in, and every degree above ambient shortens the life of the electrolytics inside. None of these appear in a peak-load calculation, and all of them are what actually arrives.
So the number that should sit at the top of the specification is not watts. It is cycles per day multiplied by operating days, and everything below it gets chosen against that figure. It changes decisions in unglamorous directions — larger margins on anything that switches, connectors chosen for service life rather than size, moving parts avoided where a solid-state option exists even when it is worse on paper.
The second discipline is assuming the failure will happen and designing the repair. Attractions are serviced in the small hours by people who did not build the thing, often in a confined space, often in the dark, often against a clock because the park opens at ten. Anything likely to fail should be reachable without dismantling something else, labelled in a way that survives a year of humidity, and replaceable as a unit rather than repaired in place.
Documentation is part of the hardware, not an appendix to it. A labelled board in a rack, a wiring diagram that matches what was installed rather than what was drawn, and a spares list with real part numbers are worth more at two in the morning than an elegant circuit. The person holding the torch has not read the design review and should not have to.
There is a version of this work that produces something impressive on opening day and a version that produces something still running in its third year. They are not the same design, and only one of them is a reference you can point at.
The commercial argument is worth making plainly, because it is usually the one that wins. A specification built around cycles costs slightly more at install and dramatically less across three years of operation, and it is the difference between an attraction that runs and an attraction that develops a reputation for being closed. Operators know this even when the procurement process is not set up to reward it.