Hacker News Reader

Comment by philwelch | Hacker News Reader

> Cost savings is always the root cause of every engineering failure, so long as you have the benefit of hindsight. I mean once you know what the problem is you can spend money to fix it, and they should have spent that money right off the bat, right?

I once heard it said that any reasonably smart person can build a bridge that doesn't collapse. It takes an engineer to build a bridge that just barely doesn't collapse, by an acceptable safety margin that still provides economy. So yes; any unforeseen failure exists because of that tradeoff.

The counterexample you name, the Space Shuttle, is just a series of fundamental mistakes compounded together--a boondoggle machine designed for a dozen different mission profiles, most of which never happened, and introduced serious regressions in survivability and launch escape compared to Apollo.

View on HN

The pithy way of saying it is "Engineering is making something with $1 that any fool could make with $2"

> It takes an engineer to build a bridge that just barely doesn't collapse

Cute saying, but an aircraft doesn't really share this property. You can make a bridge or building stronger by just adding more stuff. Many pyramids and stone bridges have held up admirably even after thousands of years.

You can't make a better aircraft by just adding more stuff. Very soon, it becomes incapable of its primary purpose.

I mean, most things will fly at least somewhat if you bolt on a big enough jet engine :) But I see your point.

The Space Shuttle disaster was a symptom of the US political process.

No, it wasn't a 'symptom' of anything. The Space Shuttle was a remarkable engineering achievement, made at enormous cost, built and managed by great people who had only the best intentions. Yet it still failed in a spectacular way, because the engineering wasn't perfect and the management wasn't perfect. In the end Challenger was destroyed by the failure of a cheap gasket, and the gasket failed because the decisions made and processes which produced the environment in which those decisions were made around making a go/no-go call was imperfect and failed. It doesn't matter what political process is in place, it doesn't matter how good the people are, or how good the engineering is, or anything else for that matter, spectacular failures are inevitable. When one happens, we have to learn from it and figure out how to build organizations and engineering artifacts that are better in practical and achievable ways, not try to use it as some kind of evidence that our pet political theory would have somehow let people make perfect decisions or make rubber gaskets invulnerable to cold weather.

It turns out that if you take some task that is incredibly hard and inherently complex and performed by a very large number of extremely qualified and smart people near the limits of human experience, and you want to do that task in a better way it isn't going to boil down to something simple or obvious.

In fact, if the US political process is so broken, why is it the only process that ends up with space shuttles sitting in hangers? It's certainly not perfect, but since nobody can point to a political process that results in better space shuttles I don't see any evidence that humans have so far developed a better political process if you are measuring it in units of 'space shuttle reliability'.

If you are not aware, I recommend you look up the design process for the space shuttle. It is a perfect example of engineering taking a backseat to politics.

TL;DR version:

The original design of the space shuttle was vastly smaller and simpler, but during the cold war the only way you could get something funded was to have a military dual use. So the Department of Defense offered NASA funding so long as they redesign the entire space shuttle to be able to capture and return Soviet satellites to the US -- a task that it never performed -- which required the design to be massively compromised.

Would you consider the Soviet Burans in Kazakhstan ‘sitting in hangers’?

The design of those shuttles lacked the flaw that contributed to the Challenger tragedy.

So, what was the engineering reason for implementing a US design requiring o-rings?

Umm... you do know the Buran only launched once with no humans aboard before being scrapped?

I’m pretty sure there were plenty of flaws that would have shown up if Buran had flown 100+ real orbital missions.

The launcher it was to use was hilariously dangerous. (Saturn V is but in a different way.) The thing would be killed before it even a achieved space. At least they managed to salvage the excellent engines out of the project.

The Russians had the good sense not to sacrifice a dozen human lives on the altar of their space shuttle program. As a direct consequence of that decision, they also still have the ability to launch manned spacecraft, unlike the United States.

Communism—which is otherwise an intolerably terrible system—led to manned space stations and safer, more reliable manned launch vehicles (namely Soyuz and Shenzhou). It’s a national embarrassment that the US sacrificed billions of dollars and a dozen human lives on the shuttle boondoggle and ended up with no human launch capacity for over a decade and counting.

Doing things that are “incredibly hard and inherently complex” and also unnecessary and dangerous is idiotic. Things like building and flying a space shuttle rather than a normal space capsule. Especially when human lives are at stake.

> It takes an engineer to build a bridge that just barely doesn't collapse, by an acceptable safety margin that still provides economy.

Holy crap, that truck in oncoming traffic just barely avoided hitting us head on!

If that driver hadn't adjusted their speed appropriately for the road conditions and planned on switching lanes a quarter mile back we'd all be goners, bro.

Reply on news.ycombinator.com