Hacker News Reader

All good points, but I do want to add context for users of this site who don't have any experience with manufacturing or physical trades - the commentary above about manufacturing welding should not be interpreted as welding being a low skill activity.

The same distinction between "manufacturing" welding and "construction" welding should probably just be applied within all types of welding in general. It's all about the application. Welding just sticks two pieces of metal together, so to use glue as an analogy - it can be used to stick macaroni to paper for a primary school activity, or used during surgery to save someone's life. Same action, two wildly different skill levels required for a successful result.

Having spent some time with prototyping shops and custom product manufacturing houses - those welders are insanely talented. There is an wild amount of knowledge required (metallurgy, chemistry, physics, electronics) to be used in practice, and the physical demands (fine motor control, holding pieces or your body under complete control at various angles/positions) should not be underestimated. Couple that with the understanding of the product being produced and how their actions can improve the end design, and you're left with what I tend to see as the closest thing to a wizard as we have in the modern era.

It's like the old saying goes - "Give a man a hammer and everything starts to look like a nail. Give a man a welder and everything starts looking like whatever the hell he wants it to be."

Came here to point out the fact that manufacturing welding, for anything that is supposed to hold against loads, is anything but unskilled labour.

I get that when it matters they will examine and x ray joints etc. But still, it really is humbling/scary to think how dependent each of us is on someone else doing their job just right. So many instances across our lifetime, one mistake and you would not be here anymore.

This dependence is a thing since someone else made your hunting spear.

Good thing, today way more than one thing needs to go wrong to end in a catastrophy (human error excluded).

One effect of increasing automation of mass produced parts is that the value of labor goes up because it could otherwise be employed operating mass production. As a result it can become uneconomical to perform the one-off tasks. Even though the assembly line and the on-site are different categories, they're part of the same economy and partially compete for some of the same labor and jobs.

So for example, in the past you might have had a machinist one off you a part. But now that same skilled operator could be programming (or feeding) a couple CNC mills that put out dozens of parts an hour. So the one off machined part is suddenly uneconomical. (Fortunately 3D printing has reduced the specialized skills needed to produce many kinds of parts, temporarily reversing the trend for some things...)

Same happens elsewhere, I assume: pipework that might have been welded on site 50 years ago is now often assembled from bolted flanged segments (which are now cheaper due to mass production) or whatever-- even though they're more likely to leak/fail over time.

Good news for the people doing the work is that they stay busy regardless. But for those who need the work done, the increased cost of one-off work can be problematic and make some kinds of project impossible to do profitably.

You are right that construction is done in artisanal, small batches. That's part of why construction is so expensive.

That's why prefabricating and manufactured buildings are such an important development: they turn artisanal into mass produced. (Alas, manufactured homes are nearly outlawed in the US. See eg https://slate.com/news-and-politics/2012/10/mobile-homes-the... or countless other articles.)

Related: this is also part of the story of why nuclear power plant construction continues to become more and more expensive, instead of cheaper.

Welding for high stress, corrosive, high vibration, high heat and high safety requirements is extremely difficult and highly skilled labour. And rightly so.

Regarding house building, there is only so much that can be pre-produced. Assembly is still on site, and that will always require rework, nobody can affors the site preparation and parts manufacture at the required tolerance levels to avoid that.

> Welding for high stress, corrosive, high vibration, high heat and high safety requirements is extremely difficult and highly skilled labour. And rightly so.

I didn't say anything to contrary. I am saying that car factories can churn out high quality welds at scale, but construction companies can't.

> Regarding house building, there is only so much that can be pre-produced. Assembly is still on site, and that will always require rework, nobody can afford the site preparation and parts manufacture at the required tolerance levels to avoid that.

Compare https://en.wikipedia.org/wiki/Manufactured_housing

It looks like 'only so much' can go pretty far.

Or modified shipping containers. Those, as well as mobile homes, have significant trade offs so.

Yes, they are trade-offs. But I'm sure many of them can be overcome, with some research and development.

Eg camper vans have also become a lot better over the years. They benefit from not being banned in the US.

I think you are unaware of what has been achieved for years already in Europe by companies like Haus - https://haus.tech/, Huf Haus, and Hanse Haus. Assembly of an entire house in 1-2 weeks implies little or no rework.

That's for the woodenframe. Excluding installation of wirong, pipes, heating... At which point we are back to manual installation, and welding where needed, of those things.

> Site or construction welding requires a lot more adaptability, e.g. ability to read plans, resolve issues when the plans are wrong, handle different materials, do layout, alignment, jigs etc and provides an endless stream of "one off" work.

Also coordination between trades, MEP systems in large spaces need lots of coordination to avoid conflicts.

I agree, it’s hard to imagine instances in which automated welding would help out pipefitters, for example. Automated welding above ceilings or inside a mechanical chase doesn’t seem like it would save much time. The welding robot would need to get onto a lift or be hauled inside the chase. A lot of the battle with pipe fitting is wrestling the large diameter pipes into place so they can be welded. Same deal with tinners, ductwork is generally in places that aren’t easy to access.

I’m much less familiar with the sort of welding that ironworkers do (95% of the construction work I oversee is remodel/working in existing spaces) but perhaps that could benefit from automation.

As a construction professional, construction automation beyond simple things like automated CMU machines seems like a very difficult problem to solve in general, in both the field and management sides.

Totally. On site welders are often more like fabricators too. Also many are specially certified for specific things like pressure vessels, water tanks, oil pipe, underwater, etc. That's all wildly different than the production line welding I've been around.

I like the concept of “category error,” thank you. I see it a lot and I am happy to have a name for it.

I agree with your points but still found the article interesting - was also a bit thrown off since "construction" more commonly applies to the site welding you mentioned rather than manufacturing/fab.