I realize this was an old thread, but it is interesting to note that the original goes back to 2015 and in that roughly two year span just how much things have changed with 3d printing. In the auto industry it is still used primarily to create prototypes and the like, but they also are using it to build jigs and the like used for conventional manufacturing, and there are companies poised to make large scale parts for the industry, if not already doing it. I read one article where Ford was talking about making some parts using 3d on a large scale, like spoilers and the like. We also have metal 3d printing now, which could make parts outside the current nylon/plastic range of things......and a lot of this has happened since the original post 2 years ago.
One thing to keep in mind is that 3d printing is a very, very young technology, 10 years ago it likely was a laboratory experiment, something Phd students work on at places like MIT (I don't know the exactly timelines, not an expert on it). Injection molding is not exactly a new technology, its first uses were back on the late 19th century, so has had roughly 130 years to evolve to where it is today, a cheap way to mass produce plastic products, so the fact that today 3d printing can do what it does is pretty amazing, and it is changing pretty fast.
Will we all be producing trains via 3d printing? I would tend to doubt it, the process of making a train like that would be pretty intense, plus there would be a lot we couldn't print and it would be costly to buy the things we can't make ourselves, like motors and the control boards. On the other hand, combining low cost/high res printers (which are inevitably going to come to the home level) with AI based design systems that don't require experience with complicated CAD design systems, will allow people at home to do more than a bit.
I will add that large scale 3d printing could potentially revolutionize the toy train industry in other ways. In theory, if it truly does what it says it can do, it would take out the cost advantage of building the trains in China so we would lose all that entails, and the other thing would be in theory build on demand might be a different experience. The kind of build technology we have today, as sophisticated as it is, still has major limitations, so it would be too expensive to build tooling for some obscure engine that people love, but in theory the combination of large scale 3d and AI could allow them to build almost anything at a price not unlike what we pay today for high end engines. Things like the motors and the control boards are pretty universal, the drive for a typical diesel model isn't going to be all that different (I am not talking necessarily to the level of a vision line engine with the steam features or the swinging bell, I am talking more a typical command control engine), and things like the sound is basically programmed in...so you could in theory order something for a road name and for a configuration and have them build it truly on demand, much the same way the auto industry used to let you order a car to order (I don't know if they still allow this, never did that myself) and they would build that model and ship it when done. It will all depend, of course, on how the 3d technology improves, how rapidly the quality goes up and prices go down, but it if holds to what is projected, what I am describing could happen and I wouldn't bet against it.
3D printing among other things compared to injection molding (again, if the technological improvements happen the way they look like they are) is a lot more flexible, much the same way that in lean production you can build 5 or 6 different models on the same assembly line whereas the old process and technology only let you build one model on each line, this is even more revolutionary than that.
About all I can guarantee at this point is that for us home users 3d printing will become better, cheaper and a lot easier to use, and that will have benefits, like creating a broken detail part or adding detail or making scenery and so forth, and likely it will go beyond prototyping they current use at Lionel and MTH et al.
