Japanese Metal Manufacturer Faked Specifications To Hundreds of Companies (jalopnik.com) 152
Last week, Kobe Steel admitted that staff fudged reports on the strength and durability of products requested by its clients -- including those from the airline industry, cars, space rockets, and Japan's bullet trains. The company estimated that four percent of aluminum and copper products shipped from September 2016 to August 2017 were falsely labelled, Automotive News reported.
But on Friday, the company's CEO, Hiroya Kawasaki, revealed the scandal has impacted about 500 companies -- doubling the initial count -- and now includes steel products, too. The practice of falsely labeling data to meet customer's specifications could date back more than 10 years, according to the Financial Times.
For rockets the concern is less serious as they generally are not built for a long lifespan, but for airplanes and cars this news could be devastating, requiring major rebuilds on many operating vehicles.
Made in Japan (Score:5, Funny)
As our Granddads believed it was.
A lesson... (Score:3, Insightful)
It's a lesson when you put unrealistic expectations on people and their performance.
They will cheat to keep their jobs.
And you have to consider how you compensate people too. Incentives can go horribly wrong. Wells Fargo is a perfect example and the financial meltdown of '08 for that matter.
And when I hear from bankers that Dood-Frank can be repealed because the problems have been addressed, I LOL. No they haven't. And it's impossible to address them. Why? Human nature.
They may have addressed the problems
Re:A lesson... (Score:5, Insightful)
Actually, I think the lesson here is that when you source material, you need to actually test to see if it meets the specifications. You should never assume that the seller is telling the truth.
Re:A lesson... (Score:4, Informative)
Testing is costly and sometimes difficult to do in-house, depending on the nature of the testing. But there is usually the option of a third party to perform the testing for you.
The other problem is you might get initial samples that are correct, but later shipments may be sub-standard. It's usually not economical to test every bit of material you purchase, especially if the testing is destructive. It should be possible to test randomly or test batches in order to detect discrepancies early rather than having questionable material in your supply chain for 10 years.
Re: (Score:3)
That's where statistics comes into play. If you study it then you will learn how many randomly selected samples to test so that you can be confident that the order meets the requirements.
Re: (Score:2)
I used to work in an iron foundry making car parts. We tested our material in house and attach our QC logs to information we gave to our customers. I'm sure some customers were satisfied with that and started building cars and tractors out of them. But we occasionally would have customers that asked questions and we'd have to show them how we do our tests (lab tour) so I think at least some of them were reading the reports.
Re: (Score:1)
Actually, I think the lesson here is that when you source material, you need to actually test to see if it meets the specifications. You should never assume that the seller is telling the truth.
Yes, the engineering / manufacturing world used to do this. The MBAs said "meh, save $".
Re: (Score:2, Interesting)
Always the MBAs fault. The customer totally wanted to spend more on the product so you could do you testing in house, but those damn MBAs, ruining everything.
Re: (Score:2)
Same thing happened with the Oakland-SF Bay Bridge. Steel was not tested as should have been required and turned out to be vulnerable to sea water.
http://www.sfgate.com/bayarea/... [sfgate.com]
Re: (Score:2)
That's right. Whenever I buy an iPhone, I destructively test it to make sure it's everything that Apple promises it to be. If it passes, I buy another one.
Of course then I have to destructively test that one, and then buy another one.
I think I personally make up about 97% of Apple's iPhone sales.
Re: A lesson... (Score:5, Insightful)
How is it "unrealistic" to expect a steel company to correctly produce steel and accurately describe its properties? Other steel companies seem to manage this without issue.
But as another responder stated, if you are a company you basically cannot trust ANYTHING outsourced these days, and must constantly monitor it for quality. Which begs the question, why outsource then if you must also incur the added cost of verification and riding herd on QA...
Re: (Score:1)
Because testing is cheaper than building a foundry (Score:4, Insightful)
> you basically cannot trust ANYTHING outsourced these days, and must constantly monitor it for quality. Which begs the question, why outsource then if you must also incur the added cost of verification and riding herd on QA
Your good options are:
1) Buy from a steel company and test a statistically appropriate number of samples
2) Build and operate your own foundry and test a statistically appropriate number of samples
You need to test either way. The question is, "which is better, buying steel from a company that is good at making steel, or build and operate your own steel company?" If you're in the business of making appliances, or bottle caps, or lawn sprinklers, or anything other than refining steel, buying from an existing steel maker is probably a better idea than launching your own foundry.
Of course there are also two wrong ways to do it:
1) Buy from a steel company and never test any of it
2) Build and operate your own foundry and never test any of it
Either of those will end up with you using sub-standard steel.
Re: (Score:2)
If you're in the business of making appliances, or bottle caps, or lawn sprinklers, or anything other than refining steel, buying from an existing steel maker is probably a better idea than launching your own foundry.
The production quantities are clearly highly relevant because some businesses are fully vertically integrated and doing very well with it. Brembo S.A. for example, they actually have their own mines, let alone foundries — and control their production chain all the way through primary distribution. But they're the world's largest manufacturer of brake parts (a lot of it just isn't stamped "Brembo" — manufacturers have to pay extra for that!) so they can keep that whole chain busy on their own.
Re: (Score:2)
Your good options are:
1) Buy from a steel company and test a statistically appropriate number of samples
2) Build and operate your own foundry and test a statistically appropriate number of samples
If the steel company is labelling its steel, it needs to test. The buyer also needs to test. You make it sound like the work is equal, but the testing is going on twice in the outsourcing case.
Of course, outsourcing + in house testing can still be cheaper. But some work is repeated due to trust issues.
Re: A lesson... (Score:5, Interesting)
How is it "unrealistic" to expect a steel company to correctly produce steel and accurately describe its properties? Other steel companies seem to manage this without issue.
But as another responder stated, if you are a company you basically cannot trust ANYTHING outsourced these days, and must constantly monitor it for quality. Which begs the question, why outsource then if you must also incur the added cost of verification and riding herd on QA...
I am not knowledgeable about auto part forgings, but for large steel rotor forgings (20-100 tons), there are about 3-5 reputable companies in the world. The equipment to manufacture such forgings costs millions, the knowledge to make such forgings is highly specialized, the cost of making a mistake and remaking the part is huge, and the volume is in the dozens or perhaps 100s of pieces per year (worldwide). It doesn't make sense for any manufacturing company to make such forgings themselves. Not even GE, Mitsubishi, or Toshiba can justify the capital and labor overhead to make such parts in house.
Obviously smaller parts are a different story, but outsourcing does make sense for most raw material forming such as casting and forging.
Re: (Score:1)
How is it "unrealistic" to expect a steel company to correctly produce steel and accurately describe its properties? Other steel companies seem to manage this without issue.
But as another responder stated, if you are a company you basically cannot trust ANYTHING outsourced these days, and must constantly monitor it for quality. Which begs the question, why outsource then if you must also incur the added cost of verification and riding herd on QA...
You never could trust something made by someone else. You always verify if material quality matters.
Re: A lesson... (Score:5, Informative)
Test?! All the time?! Millions of tons of steel?!? Every roll?!?
On September 15, 2014, U.S. Customs and Border Patrol issued HQ H209833, a ruling requiring Mill Test Certification documents accompany ALL bulk metal imports from foreign suppliers.
Domestically, what metal production facilities still exist in the U.S. do MTC testing on EVERY roll of high-carbon sheet steel, and EVERY run of high-carbon steel pipe routinely. And they have been doing so at least since the dawn of nuclear power in this country, because the NRC requires it for pipes used in the cooling systems of nuclear power plants both commercial and DOE-owned-and-operated.
I have a close friend who designed the first industrial-scale bar code printer for U.S. Steel Corporation back in the early 1980's, which they've used ever since to provide a certification paper trail for high-carbon structural steel products for a wide variety of applications. (Sadly, he was unable to patent it, because he was a U.S. Steel employee at the time, so his IP belonged to the company.) They do that both because for certain applications it's a requirement of federal law, and because it provides them with legal cover from liability in case of an accident that involves a product or structure that incorporates their steel.
So, yes: EVERY roll, EVERY time ...
Re: (Score:2)
You don't test it all, you test it by heat number. If you're receiving abuse for asking for these tests maybe you just have one of those personalities that attracts abuse. They are industry standard.
Re: (Score:2)
Re: Made in Japan (Score:2, Informative)
I believe the point was that during our granddads time the phrase "made in Japan" was synonymous with junk. Just like "made in China" today and "Made in the USA" was in the 70s and 80s... It wasn't until a good bit after after Japan rebuilt after WWII that "made in japan" started to become associated with anything resembling quality.
Re: (Score:2)
In general if there is a company or organization that is heigly trusted. Chances are there will be some abuse and corruption in time. We should always verify what we get no matter the brand. When ever we get into X is good and Y is bad then you get in trouble
Re: (Score:2)
#FTFY
Re:Made in Japan aka Jap Crap (Score:2)
My departed grandad always called anything from Japan Jap Crap. I suppose that's an apt description in this case.
Re:Made in Japan aka Jap Crap (Score:5, Interesting)
As Japan modernized and "got" how to produce quality consistently, it earned a reputation for making the best stuff in the world. Korea and Taiwan followed the same pattern, about two decades behind Japan. (Most of the world's computer components are currently produced by Korean and Taiwanese companies. Even the stuff that goes into Apple's products.) China is currently in the first stage. The U.S. probably went through the same thing after it broke off from colonial Europe.
I suspect however that there's a third stage - complacency and mediocrity. The U.S. went into this in the 1970s and 1980s, which helped Japanese products to gain a fairly sizeable foothold here. Japan seems to be going through this third stage the last couple decades, allowing Korean and Taiwanese products to eclipse Japanese as considered "best" in the world.
Re: (Score:3)
Actually, this is a common cycle. Post-WWII, Japanese stuff was considered crap [youtube.com]. The country was war-torn and rebuilding its industrial base. Most emphasis was put on volume of production and expansion, little on quality. As Japan modernized and "got" how to produce quality consistently, it earned a reputation for making the best stuff in the world. Korea and Taiwan followed the same pattern, about two decades behind Japan. (Most of the world's computer components are currently produced by Korean and Taiwanese companies. Even the stuff that goes into Apple's products.) China is currently in the first stage. The U.S. probably went through the same thing after it broke off from colonial Europe. I suspect however that there's a third stage - complacency and mediocrity. The U.S. went into this in the 1970s and 1980s, which helped Japanese products to gain a fairly sizeable foothold here. Japan seems to be going through this third stage the last couple decades, allowing Korean and Taiwanese products to eclipse Japanese as considered "best" in the world.
The fact that one company got caught doing shitty work somehow translates into the state of quality of Japanese workmanship :/
Re: (Score:3)
Takata, Mitsubishi, Toshiba, scandals just off the top of my head. Japanese workers may still be fanatically devoted to quality but management seems to be cutting corners more and more. Probably due to increasing price pressure from Korea China et. al.
Japanese quality probably reached its zenith in the late 80's to mid 90's.
Re: (Score:1)
Re: (Score:2)
As our Granddads believed it was.
"but doc, all the best stuff is made in Japan!"
Re: (Score:2)
Rather surprised AmiMoJo isn't here trying to put a positive spin on it.
Re: (Score:2)
Were Tesla and SpaceX customers? (Score:5, Insightful)
This could explain some conspicuous quality control issues in the materials, if so. There could be a huge lawsuit in the works here.
Re: (Score:2)
This could explain some conspicuous quality control issues in the materials, if so. There could be a huge lawsuit in the works here.
hUUUUUge.
But not as huge as the lawsuit from say, Boeing and Airbus, right?
Re: (Score:2)
Boeing and Airbus didn't lose a rocket due to substandard steel from a supplier, SpaceX did.
Re: (Score:2)
Wow, you fucking astroturfers really hate it when I connect the dots for people, don't you?
Re:Why trust any vendor's claims?! (Score:5, Funny)
Re: (Score:1)
There's systemd-verify which can be used to check dependency cycles and whatnot before rebooting. Other than that, you just need to ensure you're not doing stupid things. Why would you need to start and restart things? Everything in systemd has a target (where it runs) and dependencies (what is required to be started before it runs). From this it creates a tree and uses that to boot in parallel. Having things restart during boot can be an issue sometimes, I remember for example we had to reorder the network
Re: (Score:2)
Re:Why trust any vendor's claims?! (Score:4, Informative)
How much do you destroy
A few cm^3. [youtube.com]
Re: (Score:2)
Re: (Score:2)
Is this a reference to some specific quality issue with those companies' products?
SpaceX had a rocket that exploded because of a faulty metal strut, and they found several more in inventory didn't meet spec (failed at about 10% of rated spec).
If I recall, they started making those themselves after that incident.
New long-term business model (Score:5, Interesting)
Sell supplies and advertise twice the lifetime they actually have. Fold the company, let the scandal go public after investing in the company most likely to get rich fixing the problems caused by your fraud.
Double the profit, double the fun!
Perhaps insensitive, but (Score:2)
Seppuku time?
Seriously, though. Holy shit. The only way that company is going to survive in tact is if it is balls deep in some American politician's pocket that's willing to write up a bailout... Yay, capitalism.
Re:Perhaps insensitive, but (Score:5, Funny)
Seppuku time?
They tried, but the knives were made of Kobe Steel ...
Re: (Score:2)
https://www.youtube.com/watch?... [youtube.com]
Re: (Score:2)
Not saying that we shouldn't have jailed folks instead of bailing them out in the last financial crisis, but I'd also point out communism "bailed out" state companies as well.
So, the first question... (Score:5, Interesting)
How long until there is a documentable claim that this behavior killed somebody?
Next question that comes to mind: How long until I find out if my car was built with substandard materials?
Re: (Score:3)
My question: Does this change our calculus on efficacy of various alloys? If test material was equally mislabeled, there may be no safety concerns, but how we determine what metals must be present to meet certain strength and durability thresholds would likely be inaccurate.
If test material had the advertised specification but the production material deviated, then we might have serious safety concerns.
Re: (Score:2)
Surely safety concerns are large. Engineering and construction firms need to be worried. If you source from these fake stat metal guys, you probably need to put fudge factor into calculations unless the metal is tested to know what its strength really is.
Re: (Score:2)
This behaviour kills people all the time. I had heard that the Fukushima pumps didn't work because they were not actually tested (BBC report I think), but Wikipedia does not back that up. People died in Walkerton Ontario because people falsified testing data of drinking water.
"During the time of the tragedy, both Stan and Frank Koebel denied any wrongdoing and firmly held that the water at Walkerton was safe to drink. However, as the tragedy grew in severity the two were eventually part of the criminal in
Re: (Score:3)
How long until there is a documentable claim that this behavior killed somebody?
Next question that comes to mind: How long until I find out if my car was built with substandard materials?
Well, it took Takata about 10 years to fess up on their faulty airbag issues and that only happened after multiple people started dying.
Re: (Score:2)
Re: (Score:3)
> That is fraud. Whether someone died because if is irrelevant.
It's pretty relevant. If my grievance is that you sold me crappy metal, the resulting litigation will ultimately result in some type of restitution based upon loss of value directly, possible loss of market value, etc. If instead my grievance is that your crappy metal, sold to somebody else, cost me a family member, you are facing a much more open ended amount of damages- you could meaningfully be destroyed by a number of such lawsuits.
Re: (Score:2)
You're reading that wrong. It's not the difference between get over it and take action. It's the difference between huge lawsuits and possibly some jail time and even bigger lawsuits and many many years of jail time.
What's the beef? (Score:2)
Uh oh... I just hope that they have not been lying about their steak too!
Re: (Score:2)
As one who has tasted Kobe beef several times.... I think it's probably genuine. That stuff is sublime.
Re: (Score:2)
And as someone who has had Kobe beef and then had the superior beef from a certain prefecture that they don't sell outside of Japan under any circumstance, I can tell you the Kobe tastes good and really is good compared to anything out of the USA or Canada, but man, this stuff is melt-in-your-mouth. Had a very high-quality (and expensive) meal containing this particular beef, as well as fresh scallops, and it was absolutely divine, especially since I had top-shelf 23% Dassai sake paired with my meal. I know
Re: (Score:2)
Thank you - your post made me happy.
Re: (Score:2)
yea thanks (Score:2)
for last week's news
So, buildings (Score:1)
Did Kobe Steel build new headquarters at the earthquake prone area in Japan during the last 10 years by any change?
Major mishap (Score:2)
This is what we in technical terms would call a major oopsie.
Can't wait for the class action (Score:1)
Re: (Score:2)
Depends. Often parts are over engineered to account for product defect. So if this part used faulty steel it may still be good enough. As it would fail 5 years past planned end of live vs 10.
Re: (Score:2)
For a car? no. For an airplane, perhaps. If you open a jet airliner you might be entitled to a $30 rebate on your next Boeing or Airbus purchase.
Re: (Score:1)
Check your font drive shafts and sway bar for signs of premature rusting.
Re: (Score:2)
Check your font drive shafts
That's a good idea, you wouldn't want to end up with a Helvetica scenario. [youtube.com]
Screw Japanese Metal (Score:1)
Finnish Metal is much higher quality:
https://youtu.be/aNJXS9X0yY0 [youtu.be]
Re: (Score:2)
True, but not much tensile strength...
Re: (Score:3)
Compare, Japanese metal [youtube.com]
Re: (Score:2)
Now you've gone too far. Babymetal is awesome.
Re: (Score:2)
Actually, I like Babymetal. I'm just not sure it's what most people think of when they say Metal.
Materials Testing (Score:5, Informative)
Re: Materials Testing (Score:2)
Maybe most of the companies in question did test, and designed around the materials they were actually getting rather than the specs... so it may not matter in some (most?) cases.
Also could be just that there is less of a margin for failure than the design spec would indicate but it may not be a big deal depending on where the materials were used.
Re: (Score:2)
Maybe most of the companies in question did test, and designed around the materials they were actually getting rather than the specs... so it may not matter in some (most?) cases.
Doubtful, likely they'd simply refuse the batch or use it in something else that didn't require the same standard. It's pretty much common place to do materials batch testing in the first few shipments, then take random samples from future finished batches. Some companies test all batches, one of the companies I previously worked for made blades and bands for industrial saws and high-tensile lathe heads(mainly for machining medium-carbon(automotive) high carbon steel(large block diesel engines) or aluminum
Re: (Score:2, Informative)
They said 4%. It's highly unusual for manufacturers to stress-test even near that fraction of supplied parts. This could easily have been missed by the most rigorous testing regimes. These types of tests are designed to catch accidental deviations in manufacturing quality, not purposeful sabotage of the supply line.
Re: (Score:3)
Kobe Steel (Score:2)
Were their metals not properly massaged and marbled as advertised? This sounds familiar...
Already downplayed (Score:5, Informative)
The Japanese car majors are reporting "no problems" [marketwatch.com] with Kobe aluminum they've tested from the past three years. Japan Rail has said similar about undercarriage parts. There are more years, more metals and more manufacturers involved, but the pattern is clear; these issues will be pencil whipped. There is margin for error engineered into transportation products and no one is going to rip up the floor boards over paper work unless there is a demonstrative problem.
Right or wrong that's how it will be.
Re: Already downplayed (Score:1)
I work for Honda. What we purchase directly was tested and meets spec. However the impact to our supplier base is not known yet.
Re: (Score:2)
The Japanese car majors are reporting "no problems" with Kobe aluminum they've tested from the past three years. [...] There is margin for error engineered into transportation products
Not just that, but most automobiles are crash tested. There are exceptions; some exotics and even high-end luxury cars are not crash tested. For example, Audi got a lot of flak for not doing all the formal crash tests on the D3 A8, for example, but it was an iteration of the original D2 design which they did do all the crash tests on (at the time) and which aced everything. That's a slightly relevant example because it's made out of Aluminum, although not all that relevant because the Aluminum was American
If this were a Chinese company (Score:2)
I could imagine how the forum posts would generalize this event to.
Of course it's a problem for rockets (Score:2)
They are engineered to the limits of strength-to-weight, and they don't fail gracefully.
Surprised Japanese company did it (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Over there a problem is not a problem as long as you keep quiet about it.
Unlike US companies where every problem I report in management meetings magically fail to make it to the meeting minutes and every improvement I propose is met with "it's worked this way for the last 15 years, why should we change it?".
Re: (Score:2)
Nissan's stopped selling cars in Japan (Score:3)
Where's the beef? (Score:5, Interesting)
I've been hearing about Kobe on English language broadcasts for a while. One constant throughout is a complete lack of contextual information.
We all know they "Falsified data" on strength and durability of aluminum and copper... what does this mean in real terms?. Did they just check the "A-OK" box and fill in fake data without bothering to run the tests? Did they run the tests and then knowingly alter results? What is the difference between what they reported and actual conditions of materials sold? What is the risk? I would be most interested in any references that address these basic questions.
So far every downstream manufacturer who has looked into this has not been able to find anything wrong or at least they are not admitting it publically.
broadcast... complete lack of context (Score:2)
... they "Falsified data" on strength and durability of aluminum and copper... what does this mean in real terms?. Did they just check the "A-OK" box and fill in fake data without bothering to run the tests? Did they run the tests and then knowingly alter results? What is the difference between what they reported and actual conditions of materials sold? What is the risk? I would be most interested in any references that address these basic questions.
Agreed. All that the outlets I've read have let slip was: 1) there was a whisleblower, that got ignored, 2) all the numbers we've heard: 4% of aluminum sheets & rods, 200 buyers of iron powder, ongoing for up to 10 years, etc.
But what was the whisleblower's observation? Is reporting on that kind of detail just beyond the capacities of English language outlets?
My only guess is products which failed internal testing were by some process packed with good product/labeled as passing, the whisleblower had
And *that's* how... (Score:1, Funny)
Specs and their impact (Score:2)
How closely you need to monitor the specifications depends entirely on how close to the limits of the material you're designing to.
For example, if you design a bridge with a 100% safety margin (design can carry twice its rated load), a 5% variation in the quality of the steel is not as critical as it would be when the design only has a 10% safety margin.
Someone using a 10% safety margin better be testing the metal at all stages of the process. Especially if you're relying on the specific properties of a par
Not the Only ones (Score:2)
The reasons for the poor quality might be more complex than just cost saving or poor controls. There is a cultural drive in some Asian cultures towards saying "yes" when the answer ought to be "no" because they find it difficult to stand up to those they see as in authority.
Although I suspect the importers bought the stee
Discovery (Score:2)
Who posts these stories... (Score:2)
Kobe should stick to what it's best at (Score:1)
Re: (Score:2)
The stock holders are the ones who just got killed.
Yeah? So?
I hope the stock holders are able to "Claw back" the money...
Were Kobe investors some special breed of stock holder that valued quality work and integrity over share prices and dividends? "To hell with earnings! Make sure that paperwork is in order!" said no investor, ever.