Hacker News Reader

Comment by PaulHoule | Hacker News Reader

Textile manufacturing is the absolute bottom of the barrel. The town I grew up in (Manchester, NH) had the largest textile plant at the beginning of WWI but it was out of business by 1933. The industry moved first to the American South and by the time they'd paid the loans of the factories it moved again overseas.

View on HN

Textiles are something that we cannot automate with current technology. We know how to automate plastics and metals. You can buy off the shelf injection molding machines (3d printers too, but they are rarely used in production). You can buy off the shelf machines to cut and form metals. Even things like steel mills are automated with custom equipment. But sewing two pieces of cloth together is beyond current automation, and thus there is a lot of manual labor.

In the US (and Europe) manual labor is expensive, so to make something you need a lot of automation. Once it is automated the next step is enough volume to pay for the automation (which needs expensive engineers).

Of course maybe this will change. Basic textiles were one of the first things we automated 300+ years ago, but we are only able to go from a bunch of cotton to a bolt of fabric today. As I write this making a shirt (or dog bed as this story) seems beyond what we can automate. Maybe a little investment will fix that, maybe not. I'm not an automation engineer so I can't tell you how solvable the problems are.

Robots that can fold towels are an active research area

https://www.npr.org/2022/10/22/1130552239/robot-folding-laun...

towel folding doesn't seem like a step towards automating clothing, any more than building a tall ladder is a step towards ladders tall enough to reach the moon. Anyway https://ruthtillman.com/post/all-clothing-is-handmade/

I think at the root textiles just don't go where you want them to go the way, say, a rigid object or a fluid does.

Getting computer vision and physics models to correctly handle fabric is the hard part of automating what's left in textiles.

Everything else is ultimately just inclined planes and a power rod pulling on levers, which is the stuff that was solved with the Jacquard machine and a whole industry of competing models of sewing machine before we even had electricity let alone electric servos.

Robots that can fold anything might be useful in an industrial cleaning or textile manufacturing setting, but it would compete with humans who can do that in seconds. For towels, there's specialized machines (and have been for years): https://www.youtube.com/results?search_query=towel+folding+m...

There are a lot of active research areas. So far progress hasn't been happening. I don't know if it is because the problem is unsolvable (likely), or we are not investing enough into it to solve them (also likely).

> Of course maybe this will change.

It's possibly already changing? 3D-knitting (of shoe uppers is what I have noticed) can make some sewing redundant. How far that can be taken remains to be seen.

ah there are already sweaters made using this technology it seems: https://www.oliver-charles.com/pages/3d-knitting

Textile manufacturing is basically just a derivative of electricity and labor costs. There isn't much more too it, and per sq meter, it's extremely cheap to ship product around the world.

Textile mills in Bangladesh are able to pay $0.025 US cents per kwh, and factory laborers can be employed for about $150 a month. From their main port to the US west coast, when sent by container ship, costs about $0.10 per sq m. There is no universe where anyone else can compete. It's not within America's comparative advantage anymore.

It's wild to me how this movement wants to bring back all of the lowest paid, least value add, lowest skill jobs back but is totally ok with shipping highly paid, highest value add, highest skill jobs off to India. I understand that it's all a grift, the marks only care about the former jobs, and those jobs are never actually coming back, but still.

Reply on news.ycombinator.com