Slashdot Mirror
Foxconn CEO Backpedals On Planned Robot Takeover
itwbennett writes: For years now, Foxconn has been talking up plans to replace pesky humans with robot workers in its factories. Back in February, CEO Terry Gou said he expected the automation to account for 70 percent of his company's assembly line work in three years. But in the company's shareholder meeting Thursday, Gou said he had been misquoted and that "it should be that in five years, the robots will take over 30 percent of the manpower."
Foxconn executive: "We'll have a black factory in 3 years!"
Communist Party: "If you do that, all your people will be unemployed, and instead of slaving away at a phone factory, making useless crap for American idiots, we'll have people available to protest our mismanaged government"
FC: "Not my problem"
CP: "Gulag."
FC: "We'll have a 30% robot presence in 5 years!"
Foxconn is one of the most NRE allergic companies I have ever seen. Unless you, as a customer, are willing to pay the NRE for machines on their assembly line for your product, they will attempt to use the most backwards, insanity inducing flow that can be conceived of. And you will say "Hey, there's a machine for that", and they'll say "Sure, for x amount we'll do that!". And so the negotiations begin, and in the end you realize you're paying for them to build up their factory. While you will both simultaneously make profit anyway, it is entirely because the labor is so cheap, and the environmental regulations so lax, that what you're really doing is hurting your own country to make some other people very rich at the expense of just about everyone else.
Then, if you are smart, you quit your job and leave the field. If Foxconn says "no robot labor" it must mean that some major customer has decided he is not going to pay for it. The idea that the Chinese government is actually protecting its labor is asinine, but they certainly do love the press that makes it seem like it.
The problem with the notion of high automation in China is that China has a large supply of relatively cheap labor. It may not remain cheap forever but for the immediate future it will be cheap. Therefore the economics of widespread automation in a place with cheap labor become rather dicey.
Automation really only makes economic sense in a few circumstances:
1) When production volumes are high and labor is relatively expensive and capital is relatively cheap
2) When manual labor cannot achieve the requisite quality/consistency
3) When there are substantial safety issues that cannot be otherwise mitigated
Some of what Foxconn does would make sense to automate but given how inexpensive labor remains in China much automation would be terribly difficult to economically justify in many cases. My company does assembly work (wire harnesses) and even with relatively expensive US labor we have a difficult time justifying automation in a lot of cases. I could buy a machine for $1 million that would build some of our products completely automatically with just one operator required. But we would have to have production volumes in the hundreds of thousands at minimum to justify the purchase and we would need access to credit at reasonable rates. If we had the labor rates Foxconn has the production volumes would need to be in the many millions to justify. And Foxconn makes more complicated products than we do.
As an engineer who works developing flexible automation solutions, this stuff is hard and it is expensive. Sure, it is worth it for companies in North America and Europe (our main customers) because people are even more expensive. But in countries like China and India, this is more of a prestige thing than an actual business case because people are cheap and flexible solutions are not.
Now, I say flexible because the problem with industrial automation is cheap or flexible, pick one. We can easily make a machine for cranking out a product, maybe even a for a family of products. However, if it is a low demand part or worse, is not expected to be around in 10+ years, that machine will be a large useless paperweight. Those that come to us are looking for solutions for when the next product is here, they can hire an engineer to reconfigure to make it work.
My guess, this guy make a prediction without knowing the reality of actual automation and was forced to eat his words.
Slow Down Cowboy! It's been 1 hour, 47 minutes since you last successfully posted a comment
The first 3 rules are 'no puftas'.
John McAfee 'It was like that time I hired that Bangkok prostitute; to do my taxes, while I fucked my accountant'
it is entirely because the labor is so cheap, and the environmental regulations so lax, that what you're really doing is hurting your own country to make some other people very rich at the expense of just about everyone else.
Debatable. Your environmental argument is stronger than your labor argument for technical reasons, rather than reasons of rhetoric; it is the technical reasons which concern me.
Realize every cost in every product or service you buy is human labor. Profit margins come on top of human labor. This works out in aggregate, which is important for many reasons: in major aggregate, bulk purchase agreements are bound by this human labor cost and nothing else.
Consider you make steel automobile frames, which use refined steel, which is processed from coal and steel ore. Humans mine coal and steel (it's complex enough, let's ignore the cost of machines, tools, and fuel at the mining level), which mining companies sell to the refinery; humans refine steel, which the refinery sells to you.
When you buy steel, you pay a mark-up on refined steel above its material and labor costs. The refinery buys coal and steel ore as material, which the mining companies mark up above their human labor costs. In aggregate, the price of your automobile frames includes your mark-up, the steel mill mark-up, and the mining company mark-up; while the cost includes the raw human labor and nothing else. Put the mark-up and human labor together and you get the price.
GM orders 100 million automobile frames from you. For this, you negotiate with your steel refinery for 100 million units of refined steel. For this, the refinery negotiates a massive ore and coal order from the ore and coal mining companies, respectively. These large orders provide a big profit opportunity: rather than mark up $100/ton, the mining companies mark up $1/ton, selling millions of tons of material they wouldn't have normally sold, profiting millions of dollars they would never have gained if you went elsewhere. You refine the ore into steel for frames, and charge a $1 mark-up on that instead of the usual $100 mark-up, which, with 100 million frames, nets you $100 million of profit.
That's just business. You know about bulk ordering; I will explain an important factor here.
Supply-and-demand suggests this high demand will drive the prices up, yet it drives prices down. That's because supply-and-demand doesn't account for market negotiation; we shoehorn competition here. Regardless, another mechanism is at play: labor cost.
Say coal costs $200/ton to extract, because you get a 10 tonne block of coal by an hour of effort from 200 miners making $10/hr. If you try to increase supply, the next best mines--the ones not being tapped right now--supply blocks of 75% rock and dirt mixed with 25% coal, meaning 200 miners still cost $2000 to pull that block out of the ground, but it only has 2.5 tonnes of coal--it costs you $800/tonne, plus cost to sift and refine the coal out. Existing coal companies can raise prices up to about $800/tonne without worrying about a new competitor showing up with a source of cheap coal (actually higher, because they can significantly undercut and bankrupt any new competitor immediately).
Invent new tools and new mining techniques.
In this new process, the labor involved in tool production, tool maintenance, mine planning, managing, and executing tallies up to about an hour of 100 laborers averaging $10/hr to pull one 10-tonne block of coal from the ground. That means it costs $100/ton for coal.
When someone comes along and says they want 10 million tonnes of coal, you can make them a deal for $200/tonne and not go bankrupt. Below that and you run red. Discover a new process, and suddenly your competitor breaks even at $200/tonne, while you make a 50% profit margin selling at $150/tonne--a price that would bankrupt your competitor in 6 months if they tried to fill that order at that price!
You'll no
Support my political activism on Patreon.