Slashdot Mirror
New T-Shirt Sewing Robot Can Make As Many Shirts Per Hour As 17 Factory Workers (qz.com)
An anonymous reader quotes a report from Quartz: In 2015, after years of research, SoftWear Automation introduced LOWRY, a sewing robot, or sewbot, that uses machine vision to spot and adjust to distortions in the fabric. Though initially only able to make simple products, such as bath mats, the technology is now advanced enough to make whole t-shirts and much of a pair of jeans. According to the company, it also does it far faster than a human sewing line. SoftWear Automation's big selling point is that one of its robotic sewing lines can replace a conventional line of 10 workers and produce about 1,142 t-shirts in an eight-hour period, compared to just 669 for the human sewing line. Another way to look at it is that the robot, working under the guidance of a single human handler, can make as many shirts per hour as about 17 humans. The company has emerged as a leader among those trying to automate sewing, drawing the interest of businesses that make home goods and of course clothing manufacturers, including Tianyuan Garments Company, a Chinese firm that produces for brands such as Adidas and Armani. Tianyuan Garments has invested $20 million in a 100,000-square foot factory in Little Rock, Arkansas, planned to open in 2018. The factory will be staffed with 21 robotic production lines supplied by SoftWear Automation, and will be capable of making 1.2 million t-shirts a year.
So that's like... a few dozen jobs at the most? Surely with the production costs going down the shirts will be sold at lower prices, right?
#DeleteFacebook
Welcome to the future.
Low skilled jobs will truly be gone pretty damn quick. We really do need to take a look at how we will handle this as a society. If there are hundreds of thousands of people put out of work over the next decades it won't end well without a plan.
US production per worker is currently about $58,000, and seems to be going up by $10,000 per decade.
That's per capita, meaning "per person". If the per-capita output were distributed equally to every man, woman, and child everyone would have about $58.000 to spend. Each year. Including kids and babies. And they could do it again next year.
This will only go up as AI and automation take over. A huge number of driving jobs will be taken over by self-driving vehicles in the next decade (already happening with long-haul trucks), and AI and robotics will take over ever more of the production, working 24/7 and making more goods, more cheaply, and faster than humans.
We need to transition away from the current economic system real soon, or suffer massive riots and the downfall of our culture as unemployed people riot and take it down for us.
We need a way to spread the wealth out a little more evenly. UBI is one way, and we're getting really close to the point where UBI will be cheaper than the cost of government assistance plus the lost cost of higher crime and prison for the poor.
Perhaps taxing the robots and using the money to fund the rest of UBI would work.
We could also lower the SS retirement age, or go to a 4-day work week. Lots of options, many would work or could be made to work.
But we have to start transitioning just about now, or risk the downfall of our culture.
The real value of this tech is that it will allow for bespoke clothing to become a lot more accessible. Whether or not this iteration is capable of it, that future is inevitable.
This is my signature. There are many like it, but this one is mine.
I thought jails were gender-segregated. How are you going to get a broad in a jail?
Do not look into laser with remaining eye.
I honestly thought a robotic sewing machine would be called SEWER instead of LOWRY.
That name stinks.
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
I would rather actually fund education so more people would be qualified for work beyond being a meat-part in a machine, doing the same thing over and over again for days, months, years.
That was a good idea in previous decades.
Currently the number of jobs is shrinking, while the workforce is not(*). it's already causing a lot of stress in our society, and probably one reason for the recent election results.
The system was able to soak up some of the excess - the meme of children living with their parents until well into adulthood is one result - but it's starting to show signs of saturation. The burgeoning debt of education versus finding a job, currently being a topic of concern, is one bit of evidence.
Training and education are certainly important, but it doesn't address the problem. It'll only result in educated unemployed.
We need a way to support non-workers in our society, and pretty quick.
(*) Roughly speaking, population is remaining steady. Meanwhile, productivity keeps rising.
I know most people who work in technology don't really care about the "working class" who are obviously going to be affected greatly by this and other automation. It's easy to look at your DevOps-y CI/CD pipeline, see the code you write immediately go into production, and assume that it would be great to do this with factory work as well. I assure you that people will start to care when those workers start going after everyone who still has a job out of desperation, or when their jobs are eliminated through automation.
We haven't worked out a perfect solution for what happens when the vast majority of workers can't sell their labor for a reasonable price that lets them survive and consume at levels capable of sustaining businesses. Spend some time outside of the technology world, and you will see that _many_ people aren't capable of handling anything more than a job involving repetitive tasks. Lots of people need that job on an assembly line putting Part A into Slot B and adding a screw for 8 hours a day, or driving a forklift in a warehouse, or processing the same paperwork day in and day out according to rules. If you say they can't have that anymore, then you need to come up with a solution. Money's not going to disappear overnight as a store of value, and removing people's ability to earn will not end well at all. People who wouldn't dream of violence will get desperate when they lose any hope of making it.
tl;dr: Some people lost the IQ lottery, no two ways about it. You aren't going to turn a factory worker into a big data scientist. Figure out how to fix this without bloodshed, massive depopulation, gene editing/selection, Soylent Green, or similar. The things I think could work would be make-work type jobs as an employer of last resort, or just dropping the farce and giving a basic income funded by taxing means of automation. I like the idea of a basic income because along with price controls, it basically sets a floor on poverty. Let the basic income pay completely for the necessities of life, have people work part- or full-time for extra income, make it so businesses can't just raise the price of everything to compensate for the added income, and people won't have a cash-scarcity problem.
No. OMG, just No.
Instead of $12, can I get a decent shirt, made locally, not by children in Asia, for $3? Perhaps you can be employed in servicing the machines.
Please, go throw a brick through the window of a combine, or modern tractor, and insist we all go back to manual reaping and threshing. Or tear down miles of electrified fence, spill the livestock from the feedlots, and insist it's your God given right to be a shepherd.
When you realize the futility in that, then maybe you should learn to code. Or play a musical instrument. Or sing and dance. Or raise and love a child. Or extract a principle of nature from odd and surprising observations. Or recycle the mountains of plastic floating in the south Pacific, or your local landfill. When machines can do all those things then you can smash society without me getting in your way. Except, if a machine could raise and love a child, perhaps a special loving machine can be made just for you and your rage....
"Every time I see an adult on a bicycle, I no longer despair for the future of the human race." - H. G. Wells
And indeed clothing/textiles was the focus of much tech before then, the spinning wheel, the loom, the cotton gin, etc. It's no surprise it's helping drive the next stage.
There was a /. poll here some time ago about what the first technology was and it came down to clothing and the knife.
My ism, it's full of beliefs.
What do you propose we do with the people who can't just learn to code? Society doesn't pay people to do a lot of the things you mention unless they're truly exceptional. How do you propose we allow people to make a living while maintaining the money-based economy we have?
Think outside of the dev/IT world for a second. Not everyone is super-brilliant, or even latently super-brilliant. Most people need jobs that they can just show up at, perform a set of tasks, and go home when it's done. I'd argue that lots of corporate jobs paying decent salaries boil down to applying a fixed set of rules to an input stack of work. There are a lot more modern shepherds and manual farmers out there in the world than you think. Before all the factory work was offshored or moved to non-union states, low-skilled people could have a decent lifestyle. This is just the next step -- and it's not going to end well unless we figure out a balance between the Luddites and the ultra-wealthy robot owning class locking themselves in fortresses.
I wonder how big the step would be to get that robot to make a suture, sewing skin instead of a t-shirt.
Two years ago, I had a bite at a restaurant but the meal didn't sit so well. I had to go to the bathroom in the middle of the night, and for some reason lost consciousness. I got to my senses and discovered I was bleeding profusely from an actually very small cut in my forehead. I went to the doc next morning and he stitched the cut.
Two weeks later, he removed the stitches and told me he was quite happy with the result. He mentioned that the cut was actually not a straight cut, but a "hook" which apparently is difficult to cleanly close without later showing an obvious scar.
From the summary: "uses machine vision to spot and adjust to distortions in the fabric". It would be very interesting to know whether it could lay a stitch to prevent scar tissue.
8 of 13 people found this answer helpful. Did you?
You presume that the worker is at fault and not the business.
"Forget the engineers." -Carly Fiorina, briber of MIT Technology Review.
It makes sense. Once labor costs are close to zero, shipping finished products to the US probably accounts for the most significant cost. Shipping raw materials in bulk to the place where you're going to sell the finished product just makes sense.
I believe it was Henry Ford who once said that it made no sense to pay his workers so little they could never afford to buy one of his cars.
Good luck running a successful business in a world where only business OWNERS make money with which to buy stuff.
-=This sig has nothing to do with my comment. Move along now=-
For everyone saying the unemployed will rise up against the AI and automation of generic manual labor and white collar jobs just remember one thing - this advancement applies to the military as well. Once the autonomous military capabilities, including automated manufacturing, exceed the manual ones, humanity will be at the complete mercy of those few who own the armies and factories unlike any time in all human history. Good luck with that humans.
. SoftWear Automation's big selling point is that one of its robotic sewing lines can replace a conventional line of 10 workers and produce about 1,142 t-shirts in an eight-hour period, compared to just 669 for the human sewing line.
That depends ENTIRELY on the costs involved and you'll notice costs were not mentioned at all here. It's possible to automate all sorts of things but it doesn't matter if the costs don't work out favorably. Faster does not automatically equal economic efficiency. You have to do a minimum efficient volume of work and the cost of labor has to be sufficiently high to make the capital investment worthwhile. Most textile work is done overseas in countries with VERY low labor rates. It doesn't matter if it is 10 times faster if it is 20 times more expensive.
You might not be familiar with Roman (who wrote the comment you replied to), but he is a hard core religious fascist who believes he is not obligated to pay employees, ever. His next comment will likely source his favorite scripture or his personal lord and savior; he uses their statements to justify treating people like disposable commodities. He believes that business owners are evolutionarily superior to all workers and should always be treated and regarded as such.
How amusing that the captcha is "multiple", as roman also has a sock puppet account that he uses to help further the spread of his favorite parts of the gospel.
You also need to ask yourself why they chose Little Rock. hmmm. The answer is 200 miles away.
It's shameful
Won't anyone think of the children?
What do you propose we do with the people who can't just learn to code?
I *can* code and even I'm this short of moving out of it as a main occupation alltogether. Just went into a Docker introduction last evening at one of our numeros local evening dev meetups. Entire Infrastructure setup templates with 2 hours of initial scripting. Need a new instance of an entire ERP Appserver? A few clicks, go get some coffee, come back, finished.
Add in AI/Machine Learning and regular coding jobs are *over*. ... That's why I'm about to move coding as a job into a new perspective for myself. That might include getting back into dancing. That may actually have more future than my current job. And eventhough coding pays twice as much today it might be just about free in a few years.
Just sayin'.
We suffer more in our imagination than in reality. - Seneca
But there are other costs as well, like shipping.
True but this is well understood among those of us who work in manufacturing.
A lot of people would be willing to pay somewhat more for quick delivery of custom-made clothing.
While there is a market for quickly made custom clothing, it's unclear exactly how big it is and it is clear that the technology to make it happen on a large scale is not yet economically viable. I design assembly lines and production systems for a living. The economics of automating what you describe require rather substantial scale to become viable. It's kind of a chicken and egg problem. Nobody is willing to invest in the expensive automation because it isn't clear that the market exists and the market doesn't exist because nobody is willing to take the risk on building the automation for an unproven market.
I expect it will get sorted out in due time but it seems certain that either someone is going to have to take a VERY big risk or we will have to wait for the automation to become more economically viable. And it won't displace traditional mass production unless it can get very close to it on price which is a much harder thing to do than many people appreciate. It's kind of like 3D printing. It's very useful and a great technology but the unit cost of producing a single unit that way generally is higher than other mass production techniques aside from a few corner cases.