I'm sure this has been covered in the forums but it's a pretty hot topic mostly because of news media attention to "printing guns" at home. A group of modelers in the UK is attempting to get into micro-manufacturing via Kickstarter using 3D printing and high resolution scans of locomotives around the UK. It's a great idea and holds a lot of promise as well as some fundemental shifts in how we approach the hobby in the future.  The following is something I dashed off for a Facebook post so it's aimed at a non-hobby audience but thought it would be interesting to post here!  Information on the UK group is http://blog.flexiscale.co/post/44440592724/how-did-we-get-here.

3D printing is a technology that is advancing rapidly.  For those who model things (such as railroads), it promises a very exciting future.  It also will be a game changer for an industry that has had to adapt to more expensive manufacturing costs, more demanding modelers, and an aging population.  The question is, will they see that they have to adapt before it's too late?

Currently, you come up with a locomotive (or freight car, etc.) you think will appeal to the broadest group of people collecting in a specific scale. You can either make it generic enough to fit several different roads or customize it but also increasing the cost.  Then you contract with a manufacturing company overseas (they all are) to produce the model. The best most detailed ones are in brass from South Korea or Japan with the more affordable ones in plastic from other places.  Because you don't want to produce more than you need to (because excess inventory now a days can be a disaster) you make everyone pre-order so you know how many to make.  PR and sales occur and hopefully you make enough profit to keep going.

3D printing means a company could scan existing locomotives or cars (as they do in the link below) and have a very accurate representation of a real thing. You can then print one off models for each customer charging more to paint and decal.  Once you have a base model, you can then edit the CAD data to add specifics for different roads, eras, etc.  Want a supercharger on the 1982 version of the Geep you like on the UP?  Sure!  The cost to the manufacturer is the initial outlay for scanning, data storage, and the printing equipment (not cheap for industrial scale). You now have a library of products you are constantly building and have available anytime someone wants one. Sales will increase because you can make a product for someone when they want one, not hoping people will buy them. Also, your production costs drop because you don't need to ship from overseas and deal with the fluidity of foreign manufacturing. Once again, companies could say "Made in USA". And you could do all this for the price (or less) of a inexpensive plastic model, but the fine detail of a brass model which costs ten times more.

Or, if you are a group of modelers who really like that big locomotive at the museum, contract with a scanning company to grab a scan (with permission of course!) and print it yourself. Suddenly that group of modelers is a manufacturer too.  And if they are more responsive and have a better product, the big manufacturers may find themselves in a race they didn't know they were running.