Originally Posted by
HipboyScott
I think you're right about the reasonings and limiting factors... machining is not easy or intuitive at times, and despite having a lot of hours behind a manual bridgeport mill, for someone who doesn't make it their trade at least, the CNC world is still very confusing to me, yet alluring all the same. But I really don't see why a computer can't, in some time with enough well-written code in the software, figure out the optimal paths and processes for machining. It can't now, you'd get garbage. But soon enough it'll be streamlined to the point where a big market opens up that can handle some education and responsibility for the parts without having to really devote themselves to being a guru.
The target market is really destined for the makers... the people who want to produce small things for some personal or hobby use, but also want new equipment to help them start their small cottage industry, perhaps making replacement parts for old cars or niche adapters like Simon's go-pro mounts for paintball, or the Pico hopper... or make a living off of fabricating crowdfunded parts that they designed last week. These people already are out there doing what they will do, and they just want to expand their capacity with the new technology that we have today, but aren't interested in making 'CNC construction and operation' a hobby or lifestyle... just a small business. It's beyond their scope to have a shop with $30k worth of 'real' tools, they need stuff that will fit into a 2-car garage or a partial rented space in Brooklyn. There are physical limitations to how much rigidity you can get in a machine while only spending $3-5000 in metal and parts, but they can work with those limitations and the machines can too if they are designed well.
It's about the artists and designers who have been stuck outsourcing almost everything taking back the ability to make in the modern age, and they're going to do it eventually.
As you say the 3D printer thing takes off now as a 'home product' because it eliminates a lot of the factors that make it hard to make and build a CNC: rigidity requirements are minimal, the actual 'work' is very simple to a computer, there's no tool wear or part locating issues to worry about. As long as the bed is level and the part sticks, it's good.
But it must only be a matter of time when you can model up a cool new paintgun body or grip, throw it into an analysis environment of your program that breaks down the design, and figures out a streamlined method of cutting it based on a library of given bits, methods, etc that you have for your machine... and then maybe it builds up a rough body from fusing aluminum pellets into a form, before going in and finishing and boring it out to look great. Maybe as easy as swapping from a 1/8 cutter to a drill bit, a given machine can go from cutting to adding. IT also means you could machine and work surfaces that are completely inaccesible otherwise, as your bot can swap back and forth between adding and subtracting material.
And sure, maybe the user has some input during the process to choose surfaces and processes for features, clear up any confusion, and in general optimize the part to meet their design intent. But it's not like it's hard when everything is running virtually and you can track what's going on and 'figure out' how it works after a few days of screwing in the virtual environment.
Like, I don't expect it now, or even 5 years. But by around 2025 I am betting dollars to donuts I'll be able to have a reasonably priced machine (lets say $10k) in my shop that reads files from a program on my office computer that takes my crazy models from CAD and then just churns them out in materials ranging from synthetic wood to ABS to aluminum to titanium. Maybe I want to make molds for my small injection molder? No problem... I'll have 'em by tomorrow morning.
Is that consumer level? I dunno. But for people making businesses out of 'making' that's the obvious direction the movement is headed, and someone, probably a company like Makerbot, if they value relevancy, is going to invest big boy dollars into streamlining the process of fabrication for the maker.
Even if it starts with a service like Shapeways or Protomold... where banks and banks of these machines sit in formation, an army of automation pumping out random objects from all over the internet that are being designed around the world... it won't be long before that translates down to consumer level.
In fact if Elon Musk isn't pouring SpaceX and Tesla money into such technology, I'd be very surprised.
It's within enough time that we can start planning for it.