Not All the Problems are Technical...
There is much to be done before rapid manufacturing realizes its potential. It almost goes without saying that there are technical challenges, but there are political, educational and economic ones, as well. Political challenges will inevitably arise because the increased use of additive manufacturing will affect every corporate department. It will change their boundaries and possibly their relative statures. Doing that involves friction, especially when done incrementally as will most likely be the case.

The educational challenges arise in building new student curricula and in providing expensive learning tools such as software and equipment. Perhaps more challenging yet will be the re-education of a vast number of already-practicing professionals. The strict discipline acquired over years in applying tooling and design for manufacture (DFM) constraints may be difficult for some to unlearn. The situation is somewhat analogous to that presented to electrical engineers more than twenty-five years ago with the introduction of integrated microprocessors. Some couldn't quite fathom that it wasn't necessary to repair the computer if it failed, and that you could actually throw it away. But they got it, and so will their mechanical brothers and sisters. At any rate, becoming less-disciplined doesn't usually require extraordinary effort.

Economic challenges arise because of the risk associated with any change. The adoption of a new technology requires a substantial driving force. Cost is an important one, but the savings have to be considerably greater than can be obtained by, say, beating up vendor using an existing technology for a lower price. Happily, rapid manufacturing offers unique performance improvements combined with cost savings, a sum which is greater than its parts.

But a Lot of Them are...
The technical problems that face rapid manufacturing are exactly the ones that vex rapid prototyping: Speed and material throughput need to be greatly increased. It won't be necessary to match the rates of conventional processes like injection molding for most applications, and some technologies even today can produce numerous small parts in parallel at a reasonable pace. But at present it can take literally a day or more to pop out a fairly average size part of a few tens of cubic inches with good resolution. That's simply not acceptable for the majority of applications and is a serious impediment to growth. Removing this restriction will certainly lead to more applications and might even be the only specification improvement required for many of them.

Improvements are also needed in build envelope size, accuracy, finish, resolution and detail. Improving the fine performance parameters of additive fabrication is more critical for rapid manufacturing than rapid prototyping. It's one thing to lovingly hand finish a single prototype, but quite another when the quantity grows to ten or a hundred pieces. The economic benefits of RM can't be fully realized until finishing and secondary operations become manageable, and until they are, growth will be hampered.

CAD design software must be made more intuitive and must accommodate the use of complex local properties such as geometry, graded and mixed materials and color. The design software environment also becomes the production software environment making it important to vertically integrate these programs into a complete product management system.

Many more choices of materials are needed. It will take a very long time, though, before additive manufacturing methods will have access to anything like the range available to more conventional processes. Partly that's due to the long history of those methods, of course. But mainly it's because it's much more difficult to design materials which interact as closely with machinery as is required in most additive processes. It's not likely that this situation will change much in the future, either, which may well act as a governor on the speed of materials development for RM.

There are also political reasons why material choices may be slow to develop. System manufacturers are keeping this profitable revenue stream mostly to themselves through warranty policies, and by such means as electronically encoding cartridges to prevent the use of third-party materials. This discourages the very bottom-up use of the technology over a wide array of market niches that would probably result in faster growth of the whole industry. The situation is not precisely the same as two-dimensional printing where the profit is only in the ink. Three dimensional printing has potentially many more and narrower applications than two-dimensional printing. Consequently, it will require correspondingly much wider and more specialized material expertise than any system producer is likely to have. It also means that more of the profit can remain with the sale of the machinery.

Materials must also be characterized and tested much more extensively for RM than has been done in the past. Unlike prototypes or models, both short and long term properties are important to the utility of manufactured products. Answers must also be found to environmental problems such as recycling. Even today it's not easy or cheap to break a discarded product into its constituent materials for proper disposal or reprocessing. Graded or mixed materials might not be separable by any means which may result in a much lower re-use value for them. Growth of RM could be limited in strongly "green" areas like much of Europe until there are at least preliminary solutions.

A More Complicated Future...
Some have said that rapid prototyping is a solved problem. That may be true to the extent that the technology has solved the marketing problems of providing models for design verification and the other initial applications that gave rise to the field. But it is far from true that additive fabrication as a whole is a solved problem. What's far more likely is that additive fabrication - at least as it has been practiced up to now - as a solution to rapid manufacturing is intractable. It may be necessary to combine multiple technologies, both additive and subtractive, and utilize feedback control, to provide solutions that combine geometric freedom with accuracy and fast operation.

Technology combinations are appearing more frequently in intellectual property, particularly from European developers. Controlled Metal Build-up (CMB) which combines laser powder forming technology with CNC machining is an example from the Fraunhofer Institutes (Germany). Among additional examples, Trumpf (Germany) and POM-Group are working together on similar technology combinations, and Roland DGA (Japan) has proposed an additive-subtractive solution for plastic parts [9]. Purists may decry this path of development. But any farmer who can't remove a stump from his field, knows enough just to plow around it. It's not pretty, but it works.

Maybe the most important thought to take away is that a huge toolkit of additive technologies is now available. If applications can be found that are economically justifiable, systems and materials can be designed using that toolkit to suit them.