These Robots Are Using Static Electricity to Make Nikes | Hacker News Reader

These Robots Are Using Static Electricity to Make Nikes

bloomberg.com

57 pointshourislate50 comments

Their fabric handling is reasonable, but not a breakthrough. They're laying flat pieces of fabric on top of other flat pieces.[1] A human flattens out the pieces and feeds them in. This has been done before with vacuum pickers.[2] There's also an approach where the fabric is starched first to make it stiff for robot handling. After assembly, the garment is washed to remove the starch.[3]

Dealing with fabric that isn't flat is a popular robotics research topic, but results are not very good.[4] Willow Robotics used to have a demo of towel folding with a robot, but it was extremely slow.

[1] https://grabitinc.com/ [2] https://www.youtube.com/watch?v=zI7u9V5aYt4 [3] http://money.cnn.com/2016/10/11/technology/robots-garment-ma... [4] http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.689...

Just as a little fun fact -- in the semiconductor industry, this has been a major method of wafer handling for quite a while, where the wafer handling bit is known as an "electrostatic chuck." If you google that term, you can find a LOT of info on how this type of system works since it's been around, used in production and studied for quite a while!

> Miller acknowledges that shifting work towards automated factories could threaten jobs, but argues that more efficient manufacturing plants will create better jobs for displaced workers.

How exactly?

Supply Side Jesus™

Instead of the workers doing the work, they will move to maintaining the machines that do the work. Or they will move to QA to correct imperfect automation or to increase quality in general. These jobs are less physically demanding and less risky, so I'd consider them "better".

That might work for say 5% of the work force, but what happens to the other 95%. Remember, a single factory may have similar number of workers, but it's going to produce a lot more stuff so at the industry level there are going to be massive cuts when say self driving trucks become the norm.

Automation does produce a surplus, but total wealth is in many ways less important than median wealth.

Many workers will move on to other sectors and spend the money earned there to buy dramatically cheaper goods from the owners of the machines that replaced them.

I'll bet you a pair of Nike shoes that Nike shoes don't get a whole lot cheaper anytime soon.

The price will fall for an average pair of shoes available to people earning at or below the global median income, while quality simultaneously increases. As is always the case throughout the last 200 years of well studied industrial history, this process will open the market up to a lot more customers at lower income tiers to buy for the first time or to buy more. If they could afford one pair of average shoes before, they'll be able to afford two.

Nike will keep its prices high for brand purposes.

Meanwhile, robotic shoe bots churning out 3x the number of shoes in the same amount of time, at half the price per shoe that it used to cost, will be selling to the bottom 2/3 of the planet, not to Nike customers.

Robotic manufacturing will do two big things: relocate some types of manufacturing based on demand instead of labor, and make quality goods far cheaper for countries near or below the global median income wise.

those Brands like Nike will increase their bottom lines, but we all invest in these brands in our retirement accounts. Ultimately what matters is not the shuffling of money, but the increase in goods we all use. This is the creation of wealth that really matters.

You are assuming that no other sectors will be affected by automation and that most workers in that region can freely and easily switch jobs.

Programming is specifically super profitable because you need far less people to do the same amount of work. We'll have to come up with different economic models to cope with automation. One that's being talked about a lot lately is basic income, but there's various solutions to be looked at.

It won't. This is just a saying to cover their ass from political action. I say this as a robot enthusiast.

Indeed. The only relief will be the rich factory owners paying livable wages to the public via taxation.

yes and Tax consumption instead of production and it doesn't matter where the robot is set up!

We are going to have to find a way to tax robots as if they were humans doing the work. Social security can't survive in America unless we do some pre-planning. Robot displaces 10 people? Robot pays the payroll tax times 10, or something like it.

Not robots, but output of robots.

that way we just force robots out of country, we should tax consumption! that way it doesn't matter where the robots sit! and we might get some high paying jobs for maintenance back.

Where are the consumers you want to tax going to get their money from?

Government, basic income!

And where are those money coming from?

import tax, consumption tax, property tax, capital gain tax. money just needs to circulate enough!

import tax paid by whom? property tax paid by what money? capital gains tax from what?

If people don't have jobs they dont' make money only very few of them can offer to invest in stock let alone property.

So thats hardly a solution.

I am with you on basic income but it has to come from where the gains are being made and that is from those who own the robots.

I agree, my only point is if we tax robots, then they move it to somewhere else, you can't move consumption!

I am assuming people who own the robots don't work to earn money so there has to be tons of capital gain around!

increase the tax rate to %75 and you have your money for basic income. tax rate should be really high, otherwise the wealth moves very quickly to robot owners, and there should be no loop hole. that is why I think we should tax consumption because there is no way around it!

you can't have tax on consumption if people don't have anything to pay with outside of basic income. Then you just end up in the same situation that Africa is in where the economy spirals inwards.

You need access to outside value creation and that will come from the "means of production" which is the only place a society will be able to make money if we end up automating most of society.

If the people that profit from the means of production live in the same society their spending is going to be taxed as well.

my whole point is you can't tax production and expect companies to stay and pay high taxes, they are going to move to somewhere with less taxes. and if that happen and we are only relying on taxing productions then we end up with nothing! how do you suggest we set it up to prevent companies from moving? and keep in mind we have a fucked up political system and corporations own congress, so your solution can't be a complicated tax regulation because there is going to be tons of loopholes in it!

AIUI economic growth is strongly correlated with job creation. Robotics may eliminate individual jobs, but if they drive economic growth, then this should lead to more job creation that will provide opportunities for the displaced workers (though they may need education).

But why wouldn't those jobs be replaced by robots, given that we see pretty much every job being replaced by robots?

If the time it takes to mechanize a new job is shorter than a generation, the economy crumbles from the churn and permanently displaced workers. (I'd argue we're on the cusp of this.)

Everyone is addressing that we've seen displaced labor before, but not a lot about the fact that the rate at which we're displacing labor is rapidly accelerating. (Or that massively displaced labor often only gets better after a generation of massive hardships.)

Unless you're expecting "true" A.I. to show up anytime soon, only some categories of jobs can be automated, namely ones that involve fairly repetitive physical labor. Knowledge work and many types of physical jobs that require specialized expertise (such as electrician, plumber, etc) aren't going to be automated.

Oh sweet summer child, even the act of making deep learning models has been automated...

As I understand it, technology is already displacing lots of support staff around knowledge work (clerks, paralegels, clinic administrators, filing specialists, etc) and is rapidly digging at specialist knowledge (disease diagnosis, algorithmic designs, etc).

I don't expect "AI" to take over every job, I just expect technology to displace 95% of the people from 95% of the (good) jobs in ~30 years. I think once that 90% of the workforce becomes dislodged, we're not going to see the notion of "jobs" ever recover, particularly ones that are more than service industry positions.

I think we're already seeing the tip of it in lost office jobs being replaced by service industry jobs -- hilariously, replacing knowledge and skilled work with repetitive physical labor.

I'll admit that "knowledge work" is a bit ambiguous. Work where the job is to know things, sure that can be automated. Work where you have to apply that knowledge is much further away from being automated.

An interesting thought is that there might be an exodus of workers from areas that favour capital/competition to areas that support more socialist views of the labour market. Return your tired, your poor, your huddled masses...

Marx would like a word about means of production and the ultimate destination of the value they add.

Often there are government programs available for retraining if your job is lost to outsourcing or automation or similar. People who train into a new job will likely make more money, have a better quality of life, and better long-term career prospects.

So 200 people are put out of work in a factory town, but the 10 people the plant still needs are much better off. The rest of the workers are expected to take advantage of the resources available to them, and find something better.

There is still a large net job loss, but other jobs crop up elsewhere in the economy, and the general quality of jobs throughout the nation is improved. Often people don't take advantage of retraining programs, however, which leads to enormously depressed areas which used to be centered around a few large employers.

But it's not usually because of laziness or unwillingness to change gears. Many people simply balk at leaving their home community, and when everyone is suddenly out of work, there aren't openings in many industries that they might retrain into. Still, that's like staying in a village after the well dries up; sort of pointless and making life much harder on yourself than it needs to be. Automation is not a bad thing, but it requires a dynamic country where people are willing to move and train into jobs. Today's America is stagnant and stationary, a problem which would probably solve itself if these decrepit factory towns weren't so successful at garnering subsidies to maintain their lifestyles. Just look at coal country...

It somehow reminds me of http://amasci.com/weird/unusual/e-wall.html

Seems like there are better uses for this technology than making sneakers.

Could we activate electrostatic adhesion on tires to enhance road grip when braking? Can we use it for climbing walls? Walking in 0 G?

the article mentioned wall climbing as a prior pursuit, with less of a market.

Better uses? Sure. More lucrative ones? Maybe not.

SRI is working on it:

https://www.youtube.com/watch?v=I4DHfNtZGts

https://www.sri.com/sites/default/files/brochures/sri_wallcl...

Boring old rubber car tires can already grip at 1+ G. Tire technology isn't currently a safety limitation in road cars.

Electroadhesion is a cool technology. The robot pictured, however, seems to use suction generated by fans. I don't see any high-voltage electronics or wiring.

Yeah those look like brushless electric motors, the kind used in RC vehicles.

Most likely for cooling the lethal HV electronics under than protective barrier.

They have a bunch of YouTube videos. The fabric handling doesn't look like suction:

https://www.youtube.com/watch?v=_5XBNxLpBGA

This demo looks like the technology doesn't involve a lot of heavy duty electronics:

https://www.youtube.com/watch?v=8iZp-dCYp70

I'd love to see video footage of this tech in action with humans making shoes, 'demo' vids tend to be overly simplistic. Fascinating stuff!

They have wireless electroadhesion robots (drones, wall climbers, etc) so it's not something you'd notice from how a machine looks on the outside.

Case in point: http://www.hizook.com/blog/2009/08/06/electroadhesive-robot-...

Are you sure it uses fans? Usually in robotics venturi-valves are used to create the necessary sucktion to hold pieces.

This seems like a really strange approach.

https://de.wikipedia.org/wiki/Venturi-D%C3%BCse

Most over-used adjective in internet journalism - 'quietly'.

Reply on news.ycombinator.com