There’s another aspect to this reality of making things that someone pointed out for me the other day: someone’s gotta make it, but someone’s gotta pay for it too.
This thought comes from LinkedIn: I had posted a version of my “remember, it’s an airplane” philosophy on there last week and received a really insightful comment.
It highlighted how only around 10% of an aircraft’s lifetime total cost is spent on design, testing, and the initial certification. Ten percent. Think about how expensive some aircraft development programs are (we’re talking millions) and you can only imagine the costs over the entire operational lifetime.
I would imagine this estimate is more accurate for larger crewed aircraft than for the little UAVs that I work on, depending on how long each is in production, but it’s really the point of it that matters:
Design decisions you make right now, that impact manufacturability or procurement of parts or ease of maintenance, will have a ripple effect on the project as a whole.
I’m focusing on aircraft here, but you can probably extrapolate this to almost any other engineering effort.
I see three main “prongs” where this is most apparent:
- It impacts how much it costs to make or source a component of your product
- It bounds how much progress you can actually make with the budget you have
- It contributes to your customers’ opinions of, and decisions to stick with, your product
If you have some very precise tolerance on a part, that increases the time and cost to make it, each time. And if you end up needing 1000 of that one part made that is a LOT of extra time and cost.
Or if you source a super high-quality servo or other component for a use case that doesn’t really need that high of functionality or quality—is that extra cost actually getting you any return on investment?
There’s also the nuance of disposable/single-use UAVs and other products. If the component only really needs to function for a single eight-hour flight, do you really care about how repairable it is? Or what the average maintenance cycle works out to be? Maybe you can get by with something that has the physical durability required but a shorter lifespan, and is therefore a third the cost of your original choice. As long as it’s durable enough to not die as soon as it gets wet, maybe it’s good enough.
Tomorrow I’ll expand on prongs two and three. They are heavily influenced by prong one, but have their own nuances and importance too.