upvote
Toyota decided after the events of Fukushima that their supply-chain has to be way more resistant against external events. I assume they build up up to 6 months (I even remembered 18 months, but couldn't find evidence) of parts to make sure their production is not falling apart.
reply
Doesn't stockpiling parts go against some of their core principles?
reply
Just In Time only works if you can correctly forecast the lead times for parts to arrive when requested.

e.g. if you know you use 20/day and it takes 2 days to get the new parts then you order parts when your inventory hits 40.

If the error bars on the delivery dates suddenly grow, you have to start ordering parts earlier.

So in a sense, yes, needing to hold inventory is against part of JIT but if your goal is to maximize resilience to inventory shortages or supply shocks then no, it's not against their core principles.

reply
> If the error bars on the delivery dates suddenly grow, you have to start ordering parts earlier

The entire point of JIT is to identify those widening error bars early and treat. In this case, by moving more work onshore.

reply
Sure.

The error bars inherently have a big of a lag due to it taking some time for the average time to catch up to the real world conditions.

Much like trading, you could use exponential and/or weighted moving averages to give you signal as things change.

reply
Not having the parts seems like a bigger violation.
reply
A friend of mine was employed at a company which was responsible for BMW seat production. If they had issues with their production lines, BMW was out of business within a few hours/days. I guess it makes sense to stock at least the most important parts.
reply
The principles really go down to reducing waste in the overall production process.

Stockpiling parts was one method routinely used to mask instances of waste, so they had deliberately worked to reduce inventory held in stock to force waste to become visible, so they could then address the causes of the waste.

But as long as the inventory doesn’t go stale on the shelf, it is a relatively minor waste compared to the other types, and certainly much better than the major waste of having no finished goods to sell into the market.

reply
Inventory on the shelf actually isn’t waste at all if it’s needed to stabilize the system from externalities.
reply
How do you differentiate waste caused by internalities (aka something you can potentially do something about) and externalities?

It requires some deep introspection, and is often error prone.

reply
Working through that depends upon the business and the maturity of their processes, and where they are on their lean journey. There's no universal answers in lean, just guiding principles.
reply
Pendulum swings.
reply
The fukushima plant workers performed incredibly during the disaster. They harvested batteries from the parking lot and used fire engine pumps. An excellent example of anti fragility.
reply
Anti fragile systems gain from stress and instability. Fukushima is not better off from getting hit by a tsunami.

This is a great example of initiative and resilience, but a pretty textbook example of a fragile system. A single major but predictable event overwhelmed all primary and backup systems leading to a global scale nuclear catastrophe.

reply
I agree that Fulushima was fragile, but it was not a ”global scale” catastrophe. It made the news everywhere, but aside from that the effects were quite local.
reply
The logistics impact on toyota was from the area closure and evac, not anything to do with the goings on at the actual nuclear plant. Any number of disasters could have had a similar impact.
reply
None of that has anything to do with the comment you replied to. Despite their valiant efforts, it was still a disaster.
reply
It’s also not what antifragile means.
reply
I work at a Toyota dealership and I can tell you with utmost certainty that they never have anything for anything new. Usually it's on backorder and you need to hunt for a factory or distro center that has it. Then it arrives broken in the box after FedEx or UPS throw it around like a ragdoll.
reply
That's really sad to hear :(
reply
If everyone uses the same parts and components, wouldn't that put even more stress on their respective supply chains in times of crisis? I feel like it would be better to diversify parts and their production more globally, to become independent of individual geopolitical hotspots.
reply
If you have 10 suppliers all making the same wind screen wiper motor it's obviously better and cheaper than 10 suppliers all making different versions of essentially the same thing. You might argue that this will eventually lead to fewer suppliers because one will manage to undercut the others. Maybe it's on the car manufacturers to source from a wider net even if things are slightly more expensive to keep a stronger ecosystem.
reply
The parent comment is schematically correct. Right now each automaker uses its own quality standard for steel. So it’s non-trivial for a steel plant to supply materials to both Toyota and Honda.

Note that the scope of standardization is fairly limited at least for now. JAMA’s initial targets are steel, resins and wires.

reply
It's curious that the Japanese OEMs are so far behind on this.
reply
Are they? Western OEMs are jack welch asset light businesses that just assemble vehicles from tier 1 suppliers. A tier one supplier even owns the final assembly plant for one Mercedes model. All mercedes provides is the badge and mountains of paperwork.
reply
The german ones at least are very heavy on industry standardization, they have their own ISO/DIN-but-for-cars (VDA), which inspired equivalent standards now used by Chinese OEMs.
reply
[delayed]
reply
Industry standards are paperwork companies, which doesnt inherently imply they actually manufacture anything themselves. And usually standards are heavily tied into FMVSS or EPA regulations. Lots of busy work with nothing produced, but it sure employs a lot of politicians relatives.
reply
You mean the G-Wagon made by Magna Styer in Austria? If so, they are have always been made there.
reply
G-wagon has its own dedicated factory next to the ones owned by Magna.
reply
Semantics. Magna has been hand assembling / welding them them in Graz since 1979 using a network of Tier 1 suppliers. The whole engine is assembled by a single technician from start to finish, "one man one engine."
reply
Construction steel is a decent example of how this plays out. In the US almost every wide flange beam is rolled to the same spec, ASTM A992, so when I price a job I don't care which mill it comes from, and the service center can fill the order from whoever has stock. Before that became the norm around 2000 you'd see A36 and A572 Gr 50 on the same set of drawings and had to keep track of which was which. One spec didn't really concentrate supply. It made the material interchangeable, which is exactly what helps when one mill has a problem.
reply
But since there's only one supplier, they'll jack the prices way up and that will get passed to consumers.
reply
Not necessarily one supplier, but one design.

Toyota did that for years with their window switch assembly. You saw the same design on a Corolla and a Lexus, but the Lexus would have a nicer finish.

reply
But that is a single manufacturer selling to a single buyer. What is being described is a monopoly where a single manufacturer produces the parts for multiple customers. It's not the same situation.
reply
It is, but the scale is wider. If Honda adopts the Toyota window switch then they will switch to Toyota's supplier. Or Toyota will share the design with Honda's supplier and now both companies have two sources for the part.

I suspect the second scenario is the more likely, as the car makers have had long personal relationships with their suppliers and will prefer to work with them.

reply
Then one supplier buys the other and we're back to monopoly.
reply
That scenario can happen regardless if the parts are the same or not. This does not increase nor decrease the likelihood of that happening.
reply
AND… a standardized part reduces the friction against other players to compete in making the part.
reply
If the single manufacturer raises prices to increase their profits then a competitor can easily emerge and undercut them.
reply
Will they? Spinning up manufacturing of even a small part is quite expensive.
reply
deleted
reply
With JIT having evolved over time, there are usually at least two suppliers for every part and also absorption contracts (JIC / Just In Case). Same spec, dimensions, tolerances, QA parameters, etc. It insulates not only against unexpected supply chain issues but also monopoly pricing effects.

US manufacturing in the 1940s was very in tune with the multiple supplier method. Jeeps and Liberty ships still come up as examples for their broad production and interchangeable parts.

reply
> Jeeps and Liberty ships still come up as examples for their broad production and interchangeable parts.

This happened because the economy was on war footing and the government commanded it. Same reason Packard was making Rolls-Royce Merlin engines.

reply
Why on earth are you assuming this results in a single manufacturer? You have the the same number of manufacturers as before, producing the same quantity, it's just that what they're producing is according to a shared spec. Like, I dunno, AA batteries from Energizer or Duracell.
reply
Because most things in reality do not get produced that way. Even when it does happen, usually one manufacturer gradually buys all of the others.
reply
>If you have 10 suppliers all making the same wind screen wiper motor it's obviously better and cheaper than 10 suppliers all making different versions

That's a non sequitur. At least unless the whole world switches to the same standard or they all only had one customer to begin with. If those suppliers have to change, adapt or extend their production to accommodate a standard while still shipping their original product to other customers, the end result will be more expensive. And it would have done nothing to relieve the geopolitical issues brought up by the parent comment.

reply
They just might be doing very similiar things already but are just not allowed to do it that or the other way.

If you build things for A and B but they have both their own requirements and do their own tests, you might use similiar / same machines for steel etc. just switching recipes. Or your windshiledwhiper motor is the same but the casing for it is slightly different or the rating.

reply
Wait - you're saying that because they all change their product, that even though their market size jumps (they don't sell to one car manufacturer, they can sell to all) that that automagically increases costs?

Can I introduce you to peaks, troughs and shoulders?

You're likely to respond that the market stays the same, because its now split up by the wiper manufacturers that were creating specialised systems - fair.

Except nobody is ordering in an even cadence across the year, most orders will be higher at some points, and lower at others. And, the success of a given wiper manufacturer would be coupled hard to the success of the car (model[s]) they build for.

Now, the wiper manufacturers will be able to plan their output, coupled not to a given model(s) success, instead to the Japanese car industries success overall (less variable, still not deterministic though, to be fair)

That brings us to the peaks and troughs section - they dont have to have capacity for the peaks, which goes to waste (idle plant/staff) during the troughs, instead the other customers will fill the shoulders (more even output)

That's significant cost savings for the homogenised (wiper) manufacturers, with the one off cost of refitting their lines (which will be cheaper to do than normal because... multiple of them will be doing it...

reply
>even though their market size jumps

Why would it jump? Overall demand doesn't increase under the given assumptions. They'd be selling the same thing in a different form. Producing more forms costs more money.

reply
> (they don't sell to one car manufacturer, they can sell to all)

And also:

> You're likely to respond that the market stays the same, because its now split up by the wiper manufacturers that were creating specialised systems - fair.

> Except nobody is ordering in an even cadence across the year, most orders will be higher at some points, and lower at others. And, the success of a given wiper manufacturer would be coupled hard to the success of the car (model[s]) they build for.

Please try to have an honest read before misrepresenting posts.

reply
Just imagine the Apollo 13 carbon dioxide scrubber cartridge situation, but it's literally for 30k parts. Of which ~10k are ICE related.

(so, an EV without powertrain may have up to 30% less parts, or as many as 70% as an ICE. not like 10x less, at all. And this is before adding back in 7000 battery cells each made of few parts and the motor itself so... the parts count is actually the same? Interesting)

reply
A battery cell and 1000 battery cells are one 'part' from a spec perspective.

And you can also just standardize on the battery pack itself.

reply
Standardizing parts does not imply centralizing production. That may happen due to the inherent pressure to monopolize things, which companies seem to face, but they are orthogonal.
reply
If there is cause-effect relation, are they actually orthogonal?
reply
The cause effect is unrelated to standardization here. Having 2 different types of steel each of which have a single supplier is not meaningfully different than having 1 type of steel and a single supplier.
reply
Which one is easier to get, a USB-C cable or (insert esoteric cable config)?
reply
OTOH, once you have a source for "USB-C cable" / "(insert esoteric cable config)", which one's gonna be more reliable experience ordering repeatedly?

Chances are, the esoteric crap comes from one plant and has one major customer, so every bit is gonna be the same.

With USB-C cables, Who the fkkk knows what you'll get? The only bit guaranteed identical between two USB-C cables is the connector shape. Everything else is lottery.

At this point I think EU should introduce a mandatory labeling scheme for USB-C cables, so that each cable is stamped both on packaging and on the cable itself, with the very same stuff USB-C cable tester shows it. And since the "better" cables have chips on board, we could enroll them in the same "Battery Passport" / "Digital Product Passport" thing, with steep fines to all the importers or resellers who waste people's time selling fake cables.

reply
There's nothing stopping the USB-IF from fixing the problem other than laziness.

Just mandate a resistor-style colored band for the modes / versions.

The failure was not realizing that physical indication of electrical spec would be necessary in a do-everything cable/connector spec.

reply
This is a problem for consumers but a manufacturer who is trying to source USB cables for their product has no such issue.
reply
> which one's gonna be more reliable experience ordering repeatedly?

Conversely - in the esoteric world, if the manufacturer is disinclined to produce "quality" - what exactly are you going to do about it?

reply
What are you going to do about a decline in quality or an increase in prices if there is a single manufacturer?
reply
you should try this concept at home by diversifying your USB ports. Also diversify your power plugs.
reply