I'm a mechanical engineer designing products on a daily basis. We have a (really good) 3D printer and pump out amazing models all the time. Our's uses a fairly strong plastic and allows for things like moving linkages and very fine detail. The models are an amazing aid to communication and design. But ultimately they won't change our manufacturing process - the materials are too weak for our needs. So we make prototypes out of plastic and then the actual manufacturing run will be metal.
That really points out the problem with these columns that suggest 3D Printers are about to drastically change manufacturing. These 3D printers are only really a manufacturing revolution to people pumping out plastic (or weak powdered metal) products in low volumes (<1,000 units) and sizes between a few mm and meter or two cube. That actually doesn't cover a whole lot of products - the unit numbers are the thing that rules out most things you would think of.
3D Printers are great in the design phase - but a contractor like Shapeways will ultimately benefit the most from that.
Actual manufacturing of products was disrupted far more by CNC control. Truthfully a 3D Printer for all it's fancy tech is just an extension of CNC machining into building up from scratch instead of cutting down from stock.
CNC changed everything. You can CNC machine shapes that would have just been impossible before, on a size scale that no 3D printer today can match. You can use any material, and you replace REALLY SKILLED labor (a machinist) with a machine. That was dramatic. The 3D printer is just small potatoes compared to the change that has already happened over the last 30 years.
Certainly if you are designing a new phone and you want to 'feel' a bunch of cases, 3D printing is a dream come true.
But think of all the things it doesn't touch at all.
-All chemical processes, from oil refineries to pharmaceuticals
-Web processes like making paper and fabric (and even solar panels these days)
-Everything related to ICs and circuit boards
-Anything made of metal which extends from giant cranes and ships all the way down to bent sheet metal computer cases
-Anything that is high volume
-Biomedical devices and medical supplies
I think it gets a lot of press because the largest impact will be on the most visible type of product - the mass produced consumer plastic product. Today these all have to be identical copies because the fixed costs are so high (molds) and the per unit costs are so low (plastic). In the future maybe we will see more variety there, but I think fundamentally people don't like choice in products in practice as much as they do in principle (thats just my opinion, but look at what has happened with the consolidation of Linux distros over time). But I think the majority of manufacturing is not doing that stuff.
Our tool room already is much smaller than it was twenty years ago, while producing more molds, thanks to the wonders of CNC machines.
There's another problem - the authors of such articles don't understand "real manufacturing". Even when they do work with a 3D printer, they don't understand the gap between what it does and what's "real manufacturing" requires.
I think the more interesting stuff is voxel printing and some of the advanced machining technology that's being developed.
I take your central argument to be: "3D printers won't ultimately change our manufacturing process."
I take your main points to be: 1. The materials are weak and thus unsuitable for most products and 2. The efficiencies of scale of traditional manufacturing trump 3D printers in runs above 1k units.
Today your first point is true, however I use 3D printed parts as structural components in my robots and they are sound. I suspect in general material strength will increase until it meets the demands of how people want to use it. I hope you wouldn't make the argument that it can't increase in strength. There will always be material properties that can't be replicated in a fabber, but I do not know how many products in my life need these advanced properties.
Your second point ignores the decentralized nature of 3D printing. In total, yes, it will always be less efficient. But imagine a world where an Apple product launch means access to a new fabber file. Everyone in the world with a fabber could download and print the new product ... at the same time. No centralized manufacturing and distribution network could match that.
Some additional hurdles you don't mention: Material cost. $6 a cubic inch (my costs today) are really quite high. Material diversity. I need multiple materials for many objects I'd like to print. Speed. It takes hours to print something of moderate size. If I can pop over to a store and get the same item manufactured in the traditional way, I probably will.
Finally, if by any chance, someone reading this comment hasn't read "The Diamond Age" by Stephenson, please do yourself a favor and pick up a copy.
I can't imagine an Apple product that can be printed. Maybe you could print a plastic copy of the enclosure for the product, but what you'd have in your hand is a essentially a plastic design prototype. Which 3D printers are great for. But how would you print the PCB, copper traces, pcb components, battery, lcd display...etc.
So what happens when everyone is torrenting the fabber file on the launch date? Now you're pirating physical things over the net, not just data.
I think this is exactly the use case I am most excited for. Everyone can print their own objects (shared and personalized). Maybe even in their own house like desktop printing or, for more expensive equipment, just downtown like a Kinko's. Everyone can download, modify, and make unique objects with very little turnaround time and no minimum quantity.