Behind the scenes at Fujifilm's factory in Sendai, Japan

The iPhone is built in China

@ScanSpeak.

Perhaps it's not visible, due to how DPreview's reply works, but I was commenting on the original post..

@fz750

i was replying to KFC

And why can't you randomly test prior to boxing?

If we look at photo sequence 32 showing them in *large* shipping boxes, then look at photo sequence 33 saying they are unboxed...*that* is what is nonsensical...why deliberately delay the *larger* box shipment to QC one unit. You then saying the QC portion now covers the *large* boxing procedure. IMO that is nonsensical as the *large* boxing procedure can be QC'd by shipping weight, etc. In addition there is *depth* vs *breadth* of QC...are you really QCing the boxing quality and methodology, or shouldn't product QC be more about getting a perfectly functioning dust-free camera + lens, etc? Why would you want to muddle QC with large boxing...again nonsensical. Especially when photo sequence 32 shows large boxes that are wrapped in plastic sheets, etc. The context makes no sense.

That's fine, and my understanding is only via the photos and text you have provided. And if anyone thinks unwrapping and unboxing units after they go into *larger* shipping units make any sense...thats fine...respect has nothing to do with it. I await some wonderfully contrived reasoning on why this QC step matters. but the logistics of checking 1 in 10 doesn't make sense. QCing something *before* it goes into a large shipping container makes sense.

Also in order to randomly test 1 in 10 individual units actually means dismantling and unwrapping *many* large 'final' shipping boxes unless you test multiple units in 1 shipping unit which actually goes *against* randomness. So really contrived defensiveness for unsubstantiated and nonsensical claims. Do I work in actual 'physical' production and QC...no...but I work in production and QC in terms of data and software and know unsubstantiated claims (or incorrect photos/narrative) when I see one.

I guess its too much to stop being disgenious about the photo sequence where in photo 32 shows heavily wrapped up shipping parcels, and photo 33 starts with 'Well, almost ready.'. You realize there is nothing unbelievable about FujiFilm's 1 in 10 random testing in isolation of these pictures? So unless you can tell me you *personally* saw them QC from the state shown in photo 32. Isn't it just better to fix the photo narrative so Fujifilm's claim doesn't look flawed? Does it really bother you that I point out something that clearly makes no sense? We know that the photo shown in photo 33 doesn't even coincide with the narrative. This is really as simple as a Sesame Street, what doesn't jive picture sequence...

And there comes the contrived QC scenario...we are QCing for forklift drivers...oh the quality of the wood pallets must be QC'd as well...what if the weight bearing is uneven and there is weakness in one of the 2 x 4's! Lets go with occam's razor...the photo narrative is flawed and what is shown in photo 33 obviously is not the chronological successor of photo 32.

Timbukto:

This kind of 'out of the box' audit is a well-established manner to check the *whole* quality system of the factory. In this way the quality operators can check the product with the eye of the customer and this is the key here!

At this unboxing check 0 (zero, no, none) error is acceptable because in the other 90% of cases (1 out of 10, you know) the faults would be noticed by the customer.

I can understand that you may never heard about it as it is used only for products where safety and quality is first (like critical IT infrastructure or high value products - seen here ;) )

I was an engineer in such an environment for years and could not agree more with this attitude to quality!

Btw Japan is known for its good quality products anyway so all you can do is to say thanks to them carrying out all these quality steps. ;)

Yes, probably. :)

Yikes!

This is nothing new or strange. It's called a 'final random inspection' and it's always done as the very last step before shipping the products. The purpose isn't just to check the product itself, but also to check the quality of the packaging, that everything is included in the box, that the labels on the box are correct, in short that nothing has gone wrong right up until the last step before the shipment leaves the factory. This is in addition to the QC that is done before and during production (testing of raw materials, visual inspection during assembly etc.).

No this makes perfect sense and is done in two other sectors I've worked in. The benefit bottom line is a different pair of eyes may notice something else that the first pair didn't. Two pathologists here independently evaluate a specimen here and a lab manager checks one out of every 5 cases, and then on top of that a national certification rep sends in two slides that look like real cases into your batch of hospital specimens just to even check on top of that. Cost effective no , but it it's also the reason why we can rule out false positives and negatives to a 95% confidence level (and also why we correct about one misdiagnosed case a month from labs in India and Malaysia). I acknowledge two totally different industries , but the paradigm of quality control monitored by quality assurance is the same.

I'm not arguing that Fuji does not QC the final product even boxed, I'm just arguing the photo narrative is flawed and does not accurately portray the manner of which they perform this QC. They obviously should not be QCing products by pulling products out of larged plastic wrapped packages sitting on wooden pallets. Jeez you might as well have an old Japanese man pop out of my fire place when I order a Fuji product so that he can QC my unboxing experience. I am nitpicking the slightly inaccurate photo narrative not the claim of QCing 1 of 10 products after *unit* boxing. I'm specifically talking about how it makes no sense to QC 1 in 10 randomly out of large shipping parcels. Do you think any of the units at the bottom of the boxes get QC'd this way? Use your common sense. Then again most of you guys haven't figured out how Fuji ISO performance is somewhat deceptive either so there is that....

Absolutely standard and best practice to do one final statistical sample from a shipping unit - in this case, a box of lenses. In the businesses I work with (equally precision gear, different industries) the boxing process itself can cause problems. Wouldn't be a surprise if they did a 1 in 10 sample from each box, but also a 100% sample from every nth box. Your ability to know that some problem is being caused by the logistics process is based on that final sampling test.

In something with that many components, processes, and steps, and packaging as complex as that (to survive the shipping process), problems are often cumulative. A problem caused by that last boxing process may require changing torque on some adjustment screw way earlier in the process.

With smaller production runs the startup cost may be prohibitively expensive.

Meanwhile, there are business cases where the manual assembly is part of the cachet/sense-of-value of a product. An old fashioned car assembled by hand appeals to the very rich buyer far more than the old fashioned car assembled automatically.

It's quite deliberate, I am sure, that Fuji want us to see how much is done by hand. It's the assembly equivalent of putting lead weights in the body to make it feel more like a leica. They may have been compelled to use manual labour, due to startup cost of automation, but they may as well capitalise on that with such a retro camera.