JIT Vs. AM: Is Additive Manufacturing The Cure To Fragile Supply Chains?

As fascinating and frustrating as it was to watch the recent Suez canal debacle, we did so knowing that the fallout from it and the analysis of its impact would be far more interesting. Which is why this piece on the potential of additive manufacturing to mitigate supply chain risks caught our eye.

We have to admit that a first glance at the article, by [Davide Sher], tripped our nonsense detector pretty hard. After all, the piece appeared in 3D Printing Media Network, a trade publication that has a vested interest in boosting the additive manufacturing (AM) industry. We were also pretty convinced going in that, while 3D-printing is innovative and powerful, even using industrial printers it wouldn’t be able to scale up enough for print parts in the volumes needed for modern consumer products. How long would it take for even a factory full of 3D-printers to fill a container with parts that can be injection molded in their millions in China?

But as we read on, a lot of what [Davide] says makes sense. A container full of parts that doesn’t arrive exactly when they’re needed may as well never have been made, while parts that are either made on the factory floor using AM methods, or produced locally using a contract AM provider, could be worth their weight in gold. And he aptly points out the differences between this vision of on-demand manufacturing and today’s default of just-in-time manufacturing, which is extremely dependent on supply lines that we now know can be extremely fragile.

So, color us convinced, or at least persuaded. It will certainly be a while before all the economic fallout of the Suez blockage settles, and it’ll probably longer before we actually see changes meant to address the problems it revealed. But we would be surprised if this isn’t seen as an opportunity to retool some processes that have become so optimized that a gust of wind could take them down.

46 thoughts on “JIT Vs. AM: Is Additive Manufacturing The Cure To Fragile Supply Chains?

      1. Actually, that reminds me of this fascinating YouTube video from around 1910, the crew at Netherton foundry making the anchors for the Titanic by additive manufacturing. A process called “piling” where red-hot metal is forge-welded by hand on to a starter piece to make components too large to forge in other ways.
        https://youtu.be/k_LA_R4ifYk?t=319
        The whole video is well worth a watch, as is the “Part 2” one (which is the first half)

    1. Metal laser sintering folks are working on steel alloys that sinter better, and processes that increase density of the end part. Why? To make molding/casting tooling better.

      One of the coolest things I’ve seen lately are mold-sand printers for one-off castings. It’s ~10-20x cheaper than SLS, and makes sense for short runs and prototypes.

      A lot of the benefit of additive manufacturing/prototyping is still in the jigs/tooling.

  1. Make stuff (componets, subassemblies) in house? Who’d have thought….

    Don’t forget to analyse the supply chains for the raw materials though.

    Brexit is teaching the car manufacturers the same lessons. Warehouse and component manufacture in the UK. Which creates jobs locally.

  2. The idea is basically that local (additive) production provides an attractive insurance against supply chain interruptions. It might cost more 99.9% of the time, but you’re covered when ships run aground.

    The effective premium that you pay for this insurance is the difference in overseas production price minus the cost of shipping. If the cost of a metal widget is about 30x higher when it’s sintered vs cast, for instance, this premium is very expensive indeed. You’d be better off buying shipping insurance directly.

    If the idea is to spin up (local, additive) production on demand when the ship runs aground, the OP’s article describes very well why that won’t work in the last 1/3 of the piece. Parts need to be designed for additive manufacturability. If they’re not already, then the local-additive option is even more expensive b/c there’s a design (and testing?) cost as well.

    Doing all of this work within the couple weeks it takes to get the container ship unstuck is unlikely, so you’d have to have the designs sitting around already, which means doing 2x design/test work for all of the parts you want covered: once for “normal” manufacturing, and once for additive replacements.

    You’d be better off buying shipping insurance.

    1. Not all AM methods require redesign. At work we often print forgings and castings from unmodified designs using a powder-bed aluminium printer. This is useful to test part function, but not durability as the AM parts are generally stronger.

    2. There is also the economical issue of maintainng the local source that can supply the full capacity when needed. Who is going to pay for and train the workers not to mention the manufacturing equipments sitting idle? No one.

    3. Because most things need testing in their completed form for sign-off, If you’ve got 20 widgets in a device, you’d need to test 2^20 combinations of [original | AM] widgets to prove they all meet spec. That seems problematic.
      Not to mention that if some idiot tries to pull a u-turn in the Suez, all your widgets are probably delayed.

    4. >If the cost of a metal widget is about 30x higher when it’s sintered vs cast, for instance…

      I wonder, at what point will the industry realize that we have millions of unemployed people who aren’t skilled or trained enough for managerial or engineering jobs, but can be put to service in a machine shop with little additional training and can turn out simple widgets and parts far cheaper.

      You know, the same stuff that they were doing before they were displaced by even cheaper laborers in east Asia.

        1. What is the problem with employing people locally? The money comes back to you in increased demand and sales, and a lower tax burden in a society with lower social overhead to deal with poverty and crime.

          1. You’re answering your own questions. :-) The reason no one does it is because the math does not work out as you claim. Companies do what they do exactly because they already have considered all of the costs. When human labor is cheaper, they use it; when it’s not, they don’t.

  3. It seems the AM marketeer that wrote the article VASTLY underestimates the cost of having any equipment running in-house to produce parts and overestimates the cost of having parts shipped (a lot, actually probably most of which ISN’T actually JIT delivery). It takes a massive amount of space for the equipment and raw materials to do things yourself and especially small companies will have problems doing so. The scale advantage concentrated manufacturing has now given us is that a single large company with all the specialized equipment needed can deliver parts and base materials to smaller companies worldwide that can then focus on building their product. This is a cheaper solution in the long run. For many companies it’s much cheaper and easier to buy a container full of all the different components they need and just store that stock in a back room somewhere than it it so run AM manufacture when they need parts. They don’t particularly care if the parts get there in 3 weeks or 4 weeks or 5 weeks (There’s a maximum of course but it isn’t really set in stone usually)

    Indeed I question the wisdom of making plastic buckets or margarine tubs in China then shipping them across the world because (after shipping and handling) it’s .05 cents cheaper per product. But the problem is that for so many parts that tiny price difference still means that parts produced in China win out. Even if manufacturers don’t print their own parts but go to an AM manufacturing specialist in the area, the cost of doing so locally will probably be higher than waiting a few weeks and getting a few boxes full from Asia. Which is why that’s not likely to get commonplace unless there’s significant rule changes or a more lasting change to shipping availability. There’s a lot of “re-shoring” already happening, but little of it (for mass manufacture anyway) is in AM. For plastics it’s all injection moulding, because if you need 10.000 parts (or more) it’s just more efficient. For metals advances in cutting tools and machine design mean that even subtractive manufacturing can produce lots of parts very quickly and efficiently. The gains needed to make AM effective here would mean you’d need to be making hundreds of thousands of parts for it to pay off. Many manufacturers simply don’t know they’ll be making that many and are not going to make the effort.

    On top of all that, getting a design suitable for AM and actually have it run reliably for more than a single part requires a massive outlay at the start of the project in time and effort to get it “just right”. In comparison chucking a hunk of steel into a mill and getting your part out is a much simpler and more well understood process that takes much less time to get right.

        1. That boils down to the perennial issue with the Chinese: they run various export industries and contract manufacturing at a loss in order to gain foreign cash for manipulating the international market and getting western countries into deeper and deeper debt. Basically, breaking capitalism by doing the one thing that you shouldn’t logically do – refuse profit.

          1. Point being, the long term plan of the communists is to escalate the situation to a point where the Chinese people can “justifiably” call themselves the innocent victims of foreign aggression.

            You can’t deal with someone who refuses payment in kind, and the Chinese are accepting only money in trade, which is leading to a situation where the rest of the world has to stop dealing with them at some point, which then becomes the casus belli that justifies China going on the offensive and dictating world policy.

  4. “Just in time” is analog to a power supply with a small elecrolytic capacitor. It can not live well with surges.
    At first glance the solution is to increase the capacitor (local stocks) but we are dealing with money, and what is cheaper always wins.
    But money brings another complication to the probl m: Acoounting depends upon the time lapse you consider. This episode with ever given represented a huge loss along one week, but it might not be loss enough along a decade to justify the increase a conservative solution like to increase local stocks neither something as bold as switch to AM.

    Worth to mention that from an engineer perspective, globalism is not a good model as it requires all gears to be aligned to work well. And the history shows us the human beings are not capable of living in peace for a long period of time.

    On the other hand, on a humanist point of view, globalism might be persuading world peace along the last decades, and in that sense globalism is a bitter medicine.

    The lesson here is that the problem is complex, and we should not skip to any solution so fast.

    1. Until you realize the American mold you just bought is being tooled up in China. Almost all the domestic tooling has gone overseas too. I am now dealing with a ‘Tomb Cleaning Day’ delay or something on our latest fiasco from the German sounding molding venture capital group that is buying up ‘local’ molder after molder in the Western USA. What a mess.

      1. a fact-less idiom thrown out by people that Know nothing about how the world works.

        the world existed long before the Capitalism and will after it. the way we do things today is clearly self destructive but real problem is the people in power would have to give up power. and they’d rather watch the world burn then do that.

  5. No way!

    If it’s a shipping issue they could probably get the original factory to produce another batch and ship via a different route faster and cheaper than they could 3d print the parts. If the factory is just gone for some reason It probably wouldn’t take much longer to reproduce the molds and install them in some other factory than to print just the first shipment’s worth of parts. But then you would be all set for the second shipment, third and more.

    Nope, AM is great for prototypes, hobbyists and maybe a few parts with odd shapes that cannot easily be produced any other way but that’s all it will ever be.

    1. “Nope, AM is great for prototypes, hobbyists and maybe a few parts with odd shapes that cannot easily be produced any other way but that’s all it will ever be.”

      I’m so glad HaD has a search function. History would be boring without it.

  6. If you’re doing proper JIT you have safety stock somewhere that will cover the time your supplier has issues delivering. However beware the beancounters and their incessant price reductions. But they cant count it if you don’t have posession.

    1. Molds with layer lines would be the first thing that comes to my mind.
      But sure, I guess those can be machined out.

      Consider though, that mold will be stuck in some machine and ran non-stop producing many many parts per hour.
      Those parts will then flow down the production line being built up into products where the part may be buried inside somewhere, hard to service.

      If you discover a flaw was missed at some later date how much final product will need to be recalled?

      So… how quickly do you want to retool your production? Maybe a nice long testing and inspection process is a good idea.

  7. AM is “the answer” to nothing – at the very least in the near future it is condemned to remain a NICHE technology; it has its uses where it can be very handy, but it’s not displacing ANY of the standard manufacturing methods any time soon for mass produced anything. It’s goddamned 2021 FFS, being able to spot a shill should be a fucking built-in ability for anyone surfing the net by now. Otherwise I might just be a Nigerian prince with a financial conundrum that could pay handsome dividends to anyone willing to help – would you…?

  8. Injection molding is not additive. Injection molding requires tools. Additive does not.

    That being said, the nature of the parts will dictate whether it’s printable or not. Many of those imported parts are likely combinations of assembled parts. That’s something that additive does not do well right now. And when you get into assembly, you are getting into labor cost arbitrage territory.

    Don’t get me wrong, 3D printing can do a lot. But until it gets faster and more economical and more capable of multi-material, it’s not going to solve what ailed the Suez Canal.

  9. In house AM has great potential, but to me it’s looking like that potential will never be realized. The problems reside in the consumables. The classic greedy MBA syndrome drives today’s AM business model, the consumables are proprietary, frivolous patent encumbered, and hugely over-priced. Therefore AM doesn’t scale, and it probably never will.

    My desktop inkjet printer suffers from the same syndrome; I rarely use it. It’s not that the printer isn’t functionally useful, it is and I want to use it more. The problem is the ink is sooo expensive, so I only print something when it’s absolutely necessary. Yeah I can afford to buy more inkjet cartridges, but I don’t because I know I’m being ripped off. Their greed has taught me to hate the printer manufacturers. Yes I’ve tried refilling spoof-chipped cartridges and using Chinese ink tank kits, but the results have been unsatisfactory so far. Too messy, unreliable, inconsistent print quality.

    1. “the consumables are proprietary, frivolous patent encumbered”

      Are you talking about filament?
      Or are there patents on the metal powders used in metal printing?

      I don’t think AM has any potential to realize when it comes to mass producing plastic parts. But that’s due to speed, energy efficiency and quality, not consumables. If someone ever did build a 3d printer with a factory production line in mind no doubt it would use a pellet extruder, not filament. It would consume the same stuff the typical injection mold machine does today. Have you never seen one of those giant printers that people make life-size statues and usable chairs with? That’s how those work.

      But it would still be slow, wasteful of energy and make parts that are prone to delamination.

      On the other hand for prototyping, one-offs, non-moldable shapes, hobbyist and educational use AM not only has huge potential but I cannot see how you could say that potential isn’t being realized. It’s everywhere!

      I don’t know as much about the state of metal printing if that’s what you were talking about. But patents do expire. Sometimes it gets drawn out with extensions and staggering patents with trade secrets. This is one community though that will take notice and make use of every idea as soon as it hits the public domain.

  10. Just saying JIT does not mean what you think it mean..
    It doesn’t mean to make parts in china. AM is very more JIT/one piece flow than making batch in China. JIT as a bad reputation because people use the terms incorrectly.

Leave a Reply to JohnCancel reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.