Hacker News Reader

Comment by chrisfosterelli | Hacker News Reader

The article appears to use these individual cases to draw the conclusion that "speed and competitive pressure in the private space race make workplaces necessarily unsafe". While tragic, I don't know how a few individual cases can show that.

Is there any actual statistics that support that claim? The article says that at least 24 workers have been killed while in the space industry since 1980. Compare that to coal fatalities, of which there were nearly 2,000 fatalities over the same period.

SpaceX has 12,000 people and are making rockets so it seems like some degree of danger will be inherent. Of course this should be reduced where-ever possible, but if the claim is that the space industry is unnecessarily unsafe then I'm not sure that's convincing.

View on HN

While I don't have numbers ready to cite, the steel industry is a superb analogue. It, too, is a hazardous production environment. It, too, has production rate as an OKR. I believe deaths in that industry used to measure in the thousands [1].

There's a history, at least in the US, of production rate correlating to fatality rate in certain hazardous industries. That said, one of the most famous steel CEOs (Paul O’Neill at Alcoa) put safety at the forefront and transformed the company and their production [2].

1: citation needed

2: https://www.forbes.com/sites/roddwagner/2019/01/22/have-we-l...

But the steel industry has been doing the same thing (with largely the same facilities) for the last 50 years. The move to recycling and mini-mills is the biggest change in the last 40 years. The same is largely true of aluminum mills. You should be able to figure out how to make that safe, but you won't make anything, if you demand that everything be as safe the first time as it will be in industries, which are a century old.

Aerospace industry is 120 years old.

Well, it was certainly a lot more dangerous then:

https://youtu.be/ZNAHcMMOHE8?t=395 : A trip to the moon 1901

But seriously, the point is that the techniques and facilities are not changing, which is blatantly not true of reusable meth-ox rockets. Or you can just say we'll stick to Saturn Vs to get to space.

Your point is moot

Are you referring to the infamous Wright Brothers rocket ship?

By that logic the steel industry is 4000 years old.

With occupational safety, just like anything else, you can hire professionals and set whatever objectives and deadlines your budget allows. Companies that embrace it get much better and faster results than companies that resist it.

Your argument doesn’t hold well as the coal industry has been following the same recipe.

>While I don't have numbers ready to cite, the steel industry is a superb analogue. It, too, is a hazardous production environment. It, too, has production rate as an OKR. I believe deaths in that industry used to measure in the thousands [1].

I recently read RAILWAY ADVENTURES AND ANECDOTES (1885) <https://www.gutenberg.org/files/31395/31395-h/31395-h.htm>, a collection of interesting moments from the first 50 years of railroads. Almost every page there is a mention of some fatal accident for passengers and crews (usually a boiler explosion, and/or a derailment), yet everyone seems very matter-of-fact about such things despite (or, rather, because of) their frequency.

I think the rough part of that analogy is that steel is at a different stage in its lifecycle as a product. Some innovation, to be sure, but nothing like this pace. Here, iteration speed is paramount.

Perhaps SpaceX should move somewhat more slowly and focus more on safety, but I hope that people keep in mind that the only way to be perfectly safe is to do nothing with any risk, which in this case translates to making no real progress on rocketry. That comes with species-level risks.

And if you double the amount of stuff done for each iteration, it might increase production time by much more than double, or potentially make some of the more difficult things impossible, as focus and momentum is lost, and mental state decays more between iterations.

When my father worked for Republic Steel (1958-1969) it had a very strong safety culture.

The wording of that sentence threw me off, he wrote 'necessarily' but my mind heard unnecessarily as well. I think the gist of the article is more to bring attention to the technicians who carry out the dangerous work than to assign blame to the industry. I think the author meant, it's dangerous work, but necessary to get it done quickly to be competitive.

Well, I interpreted that a little bit differently. The sentence was:

"speed and competitive pressure in the private space race make workplaces necessarily unsafe"

I took that as meaning "private space companies are moving too fast and cutting too many corners to keep their workplaces safe". Meaning the "necessarily" part is only because companies are choosing this speed (or, similarly, being forced to move at this speed because other companies are also moving too fast/cutting corners).

In other words, I think the author's intent is more damning than you are implying.

Right, I meant necessary from the companies perspective. The author was certainly bringing to question the necessity of the industry from the broader perspective, his reference to Gil Scott Heron's song couldn't have made that clearer.

This is surely right. You wouldn’t use “necessarily” here to mean, like, “justifiably”. It means that it is a direct consequence of “moving too fast and cutting too many corners”, as opposed to an orthogonal safety lapse that can be shored up without changing the underlying circumstances. Inexorably.

Not to say the article is correct about this, of course.

I'm not crazy about the way the author is using "speed" here. The word is not only doing a lot of heavy-lifting, it is ambiguous because rockets can't be slow. "Rushing" or "hurrying" would be less ambiguous yet not as peremptory. I think using "haste" would be unambiguous yet still transmit some imperativeness.

A sample size of one is definite proof for anything I already think is true!

Anecdote or spot data is definitely evidence though!

5 year Osha reported fatalities & serious injuries:

https://www.osha.gov/pls/imis/accidentsearch.search?acc_keyw...

Osha incident report for this incident:

https://www.osha.gov/pls/imis/establishment.inspection_detai...

Also had they followed established safety standards the incident focused on would not have happened.

That actually sounds like a mature and potentially safe industry.

This is an important point. We have regulators already working in this space, and OSHA and the like actually do a great job. When people mess up (and they do!) there's an existing bureaucracy to step in and force the companies to correct their processes. So it's entirely reasonable to feel that SpaceX is on the whole a safe industrial environment and that their OSHA fines were reasonable and just.

I think that’s fair. It’s worth noting that there are no shortage of NASA-related technician deaths. SpaceX already has 12,000, overall more employees than NASA and contractors at KSC and Cape Canaveral, and yet has no fatal incidents. That’s evidence to me that SpaceX, while it can improve more, has no worse safety than has been common overall in this industry. I think claims about how the industry is going “too fast” because they’re private miss the mark. SpaceX does some things that improve safety over the status quo, such as automating testing. Using automated stacking of the Starship stages (instead of labor intensive manual stacking) is potentially a big improvement.

Consider that there have been about 9 or 10 worker fatalities due to the Shuttle program (not counting the astronauts) in 135 flights of Shuttle vs zero SpaceX fatalities in nearly 200 Falcon launches, and I think it’s clear that finger wagging about private space “going too fast” is missing the mark. It’s actually possible that the slow practices of traditional contractors contributed to some of these deaths whereas the modern automation-driven SpaceX approach has limited fatalities.

https://en.wikipedia.org/wiki/List_of_spaceflight-related_ac...

> Also had they followed established safety standards the incident focused on would not have happened.

Exactly.

I like to dump on Musk, but people disobey safety regulations all the time, and sometimes it bites them:

"Three Dead After Falling Into Molten Steel Pit" https://apnews.com/3c42c98020afcd458f1f49a861d0d87b

In this instance, the main culprit was the fact that they didn't dry out the mold properly which then cracked and leaked molten steel everywhere. However, people were not supposed to be standing on top of the cover plates during centrifugal caster operation.

There is a reason why you obey the safety regulations every single time.

I don't know what the figures are for agricultural accidents in the US but in the UK they're roughly 20 times the fatality rate of *every other industry in the country combined*, with a pretty steady 40 deaths per year since 1980.

If the US accident rate is similar (let's be generous - I suspect it's worse) then a rough back-of-an-envelope calculation puts that at 8000 deaths per year, and *you don't even hear about that*.

The measure for safety shouldn't be the number of employees or the number of casualties but how preventable an accident is and what is the reason why it could not be avoided.

That's the point of the OSHA investigation but number of casualties is a comparable metric in a way that "how preventable an accident is" is not.

Well, if someone refused to follow an established standard, that's preventable.

Nothing about this is because they're private. The faster you innovate, the more frequently you see accidents. For decades space programs were almost entirely government work and they suffered their fair share of accidents (in some cases like the soviets, considerably more). Conversely aviation is almost exclusively privately funded and one of the safest ways to travel is by air

> The faster you innovate, the more frequently you see accidents.

Do you though? I mean, again this article only cites three, and if that's all there were that seems to me to be a very reasonable accident rate in heavy industrial environments. People get crushed by backhoes almost every week just digging foundations and paving roads!

I imagine there are at least two orders of magnitude more construction workers (and backhoes) than rocket builders (and rockets).

At the rate SpaceX is hurling rockets up, that might not hold true for long! ;-O

aviation is not exactly a laissez-faire industry with private industry self-regulating...

In fact, government programs like Shuttle had MORE fatalities (9 or 10 technician fatalities, although you can quibble with some of them) in fewer launches than Falcon/SpaceX (none).

https://en.wikipedia.org/wiki/List_of_spaceflight-related_ac...

Plus 14 astronauts, at least half of which were because the government management overruled engineering because they wanted the PR of a successful launch (and didn't get it).

Haven't we had more spaceflight related deaths than the Soviet Union + Russia?

And 10x the amount of spaceflight?

No I'm pretty sure they have run more missions too. Remember after the shuttle and before SpaceX they were the sole ferry to the ISS.

There was like 1-2 flights with Russians per year, SpaceX flies much more.

https://en.wikipedia.org/wiki/Nedelin_catastrophe at least 54 deaths http://news.bbc.co.uk/2/hi/world/monitoring/media_reports/70... at least 48 deaths

> if the claim is that the space industry is unnecessarily unsafe then I'm not sure that's convincing.

The claim is that it's necessarily unsafe. You even quoted that part.

It is worded funny in the quote. In the context of the entire article, the author is clearly saying that the "speed and and competitive pressure" are the casual factor for it becoming unsafe (i.e., given the requirement of going fast by higher ups, it is necessary to be unsafe).

So he's implying that if they weren't so focused on speed and competitive pressure it could be more safe. There's where I get "unnecessarily unsafe" from, because in the author's perspective it's not necessary to prioritize speed and competition over safety. My point is that I don't think it's obviously unsafe at all, or at least could use some statistics to support his argument that it is.

There will always be risk/development tradeoff. You can play with the balance but the only way to bring it to zero is to stop development.

That's not to say that improvements can't be made. The real question is how many of the injuries and deaths were preventable by following the guidelines that already exist, and are generally followed but not perfectly followed.

In the main example sounds like they didn't follow the safety rules that already exist / should have been followed.

I agree. If the majority of your business and industry is following the safety standards, then it shouldn't cost much time to increase compliance in a rare few occasions.

This is why I don't think the argument that development is too fast holds water.

On the other hand, if the company were flagrantly ignoring safety standards in their process, yes, it would slow it down.

Ah, I see, that makes sense.

I agree with the author that it shouldn't be necessary to prioritize speed and competition over safety. The whole "will we get to X before country Y does" is IMO childish, dick-measuring nonsense, especially if getting there faster means more people have to die... because in the grand scheme of things, it doesn't matter all that much if humans "get to X" this year or next year or 5 years from now.

At the within-one-country level, you can eliminate concerns about one company out-competing another (via unsafe practices) with regulation. Globally, sure, you can't mandate that another country slows down in order to be more safe, but I think the downsides of allowing another country to get there faster are overblown.

If we look back at the US-Soviet space race during the Cold War, it doesn't seem all that clear to me that the US benefited all that much from being first to land on the moon. Consider that the Soviets won the first milestone in that race, and that many people don't even know that or have forgotten. And consider that, post-moon-landing, the space race was essentially over, and both the US and Soviets went on to focus more on military/intelligence and commercial applications for the ability to get things into Earth's orbit, and stay there. Sure, there were more moon landings after the first one, but I think while most people know who Neil Armstrong is, they couldn't name any of the people who landed on the moon in later missions. Hell, most people don't even remember that Collins flew with Armstrong, and I imagine there are quite a few who don't remember Aldrin landing as well on that trip. (How many people would actually recognize the name Gene Cernan, the last human's feet to touch the moon's surface? I hadn't heard of him until I just looked it up.)

Yes, you can make the argument that the US got some big-time global bragging rights for making it first, and that certainly conferred some economic benefits, but I would argue that the US's economic prosperity after that point, as well as the Soviets' eventual decline, was pretty much baked in at that point, and would have happened regardless of what happened with the space race.

So this is just a long-winded way for me to say that I don't think "getting there first" matters all that much, and that we should slow down and make sure all of this is safer. Sure, we're never going to eliminate all risk, and sending humans into space will number among the riskiest things we do for quite a while into the future, but there are levels of safety and unsafety that we should and shouldn't accept.

I agree that it likely is a nothingburger, but I wonder what the ratio looks like. The coal industry was probably much more sizeable than the number of workers working on rockets.

> 24 workers have been killed while in the space industry

Non-competitive US Government incompetence killed 14 astronauts since 1980.

they were fined by OSHA for severe safety violations, and they are also compared to a competing space company that is doing much better on safety. The incident does not warrant apologism of SpaceX or any "whataboutist" comparisons to other industries.

It's possible to both trust OSHA's findings and disagree with the article's assertion that the space industry in general is unsafe.

it's a funny sentence in the article everyone is picking apart, of course the space industry is not necessarily unsafe, the point seemed to be, and I may be misreading it, that the space industry happens to be more unsafe than it needs to be due to the competitive incentives at play.

Yes, I agree that's the point, and that's the point I disagree with. I don't think that the article contains robust evidence that the space industry as a whole happens to be more unsafe than it needs to be. It focuses on a single case study and cites what would generally be small accident numbers in any other industry. While this case is tragic, I think more evidence is necessary to establish the trend that the article is attempting to show.

Reply on news.ycombinator.com