So over the next few years I expect there will be quantum shifts in the way personal manufacture works, and along with them Thingiverse’s content will expand in character.  Already the things you see on Thingiverse usually are no-support printable, are virtually always manifold, and are generally scaled properly.  This is now the ground state.  I have not forgotten that it was not always this way.  I uploaded this without irony, after all:

Woefully unprintable (never was printed, either), won’t interface properly to any motor around, and I was one of the responsible ones!  Times have changed, though, now it’s almost expected that you at least offer some advice to get things printed, and on many if not most of the new things, you can find at least one picture of the thing already made, which are often gorgeous:

But what’s coming is new technologies.  What’s coming are the powder printers, and beyond them, the deeper leveraged systems that get us access to the realms of the very small and the very large builds.  There are designs for bigger CNC tables here already, as well as DIYBio to get us access to the very small world of genetics.  The infrastructure of Thingiverse isn’t intrinsically limited to squirting or cutting plastic with nozzles and lasers, it’s general, and it can handle powder-native designs like this one easily:

And when things start showing up that are really wild, it’ll handle those too.  (Although of course, a flexible model that harkens to the shape of C-60 is pretty wild if you ask me…)

Super-late post on the super-neat option of bezier surfaces in OpenSCAD here– William Adams is extending OpenSCAD to create a whole new set of options for modeling in text!  Bezier surfaces are one of the fundamental (if less well-known) primitives of 3D modeling, so this is definitely about more than being able to make wavy surfaces (although the wavy surfaces will doubtless be many and beautiful).  Sweet!

So I finally got my browser properly updated and messed around with TinkerCAD.  TinkerCAD is a new in-the-browser 3D tool that lives in WebGL and spits manifold STL files natively.  It’s an interesting middle ground between 3DTin’s voxel system and more advanced software like CloudSCAD, and by the way did you notice how all three of those are web-based browser modelers?

The Cloud is here to stay, certainly, and it seems to be good for the diversity of 3D modelers.  (Of course, the availability of cheap 3D printers hasn’t hurt either!)  There are still hurdles, bottlenecks, and so on with modeling in the cloud, but it certainly is vibrant lately!  The native ease of sharing on these platforms is definitely something to watch.

But in practice, many of the awesome things on BlendSwap are very non-manifold, broken into many objects in a way that messes with .stl exporting, or any number of other things to confound the would-be 3D printer.  This is not to say you won’t find quite a few printable goodies on BlendSwap, but right now, a lot of them won’t be printable unless you’re willing to sit down with them and do some pretty rough topology work.

Of course, this should not stay this way.

To change this, all we really need is these two things:

STL Omnivore: A script in Blender that seeks out all “traceable” polygons in a selection and pushes them to the same STL.  Ideally rolled right into the STL export for Blender, but it’s a thought…

Solidify: A better way to get slicers to make up their minds about what the inside of a non-manifold object is.  Holes are everywhere.  Thin shells are a way of life in the animation community.  These are mathematical impossibilities in the real world, but 99% of the time, an animator is after something that at least could exist in the real world.  In the long term, our slicers should be making better and better guesses at what they meant to do.

(There are already quite a few tools to “clean up” meshes, so what I’m really thinking of is better automation, shorter toolpaths here…)

We focus on solid CAD in the 3D printing scene a lot because we early adopters are more mechanically inclined, but I think within a year or two the field is going to seriously swell with CGI artists, many of whom will be bringing an entirely different tools background with them.  As they arrive, I think those of us who really dig in and start making tools for artists have a chance to score some big open source karma points making the above tools, and while I’m on the subject, here are some other goodies that’d be nice:

Bump mapping: I mentioned this a while back when Adam Meyer’s amazing lampshade script showed up, and since then I’ve done a little bit of research but haven’t been able to even consider rolling it into something that could print.  At some point though, someone is going to make a slicer that reads the UV coordinates of the polygons in a mesh, and then we’ll be able to bump-paint models WAY above the resolution of the mesh…

Image mapping: dual and multiple extruder modules are currently high on the wish list of RepRappers and MakerBotters, and not just for support material.  Imagine three colors of ABS patterning down at small scales to make colored models, or powder printers that make use of UV data to do colors.  This tech exists in the closed-source 3D printing world, but the principles don’t seem too intractable to me…

We’re in an amazing time with a super-deep set of amazing tools that are setting people’s imaginations loose in a mechanical wonderland of gears and springs and latches and snaps and buttons and bolts, but the next big thing in tools for 3D printing might be the stuff that lets the people who’ve been making hopping lamps and singing frogs and walking water towers print with a rapidly shrinking amount of fuss.

Seems like there are a lot of videos and whatnot of various applications formerly thought of as “desktop things” being done on the new tablet form factor. There’s a fundamental problem of interface which hasn’t really been fully solved yet: 3D is more complicated than 2D, and even when this was on the desktop, your interactions are limited to a 2D screen.

Solving this interface problem has been helped along by the existing tools: mice and keyboards, which are, despite having form factors that bulge out at you in all three dimensions, are still basically 2D interaction tools. Like in this example: the user is fiddling with a cursor using the touch interface, which then allows 3D interactions.

It definitely looks a little awkward at this stage. I think to really take off, a multitouch interface is going to have to do some serious scrapping of old pointer paradigms and get into what will at first seem pretty surreal. The Blender Conference last year had a demo where a tablet’s multitouch was used as a kind of digital puppetry rig for animation, so it’s definitely being thought about. In terms of more “natural media” interfaces like sculpting, there’s certainly some obvious advantages and ways the natural motion of the hand could be incorporated in ways which, to the user at least, would feel as natural as sculpting clay, but there’s not much call for, say, hand-sculpted mounting brackets.

There’s also a bit of a problem with real estate. On the tablet form factor, screen real estate is at a premium, which isn’t helped by having the user’s opaque, starfish-like appendages blocking the screen. Maybe there’s something to the claims we’re getting from Microsoft lately that future interfaces will basically be whole rooms…

Open Source Hardware is a big idea. Arguably it’s THE big idea. We implemented a system of value-capture for ideas during that brief period where an infrastructure for replication of ideas existed, but was costly. Now with the cost of replication plummeting this system has, I would argue, already become mostly obsolete. Those open source farm machines? Way of the future right there.

Okay, found a really sweet Open Source Hardware project that should serve as an inspiration to Thingiverse users with a hand in the electronics part bin, the Seeed Studio DSO Quad.  Not only is this the second stage in the major shakeup to the way scopes work that was started back with the one-channel wonder that popped a couple years ago, but the firmware and hardware are open source!

I think there’s a new mindset starting to take hold, at least in the upstart companies that make developer tools (like 3D printers and scopes), that being open makes business sense.  Sleek-looking products don’t have to be closed, and I encourage Seeed Studio to consider uploading any solidworks files they might have of the case to Thingiverse where doubtless we could make a whole heap of groovy accessories for it.

Neri Oxman is working on foundational technologies for 3D printed architecture lately, taking advantage of the key difference between CNC and conventional manufacture: complexity is decoupled from cost.

Now anyone who’s tried to print a particularly hairy model on a thermoplastic printer knows that this isn’t perfectly accurate– complexity often adds little wrinkles to the build, but the curves are definitely shallower.  A single-pour concrete structure with the air ducts built right in is a ridiculous idea, but a 3D printed concrete structure… that can be done.  Oxman also has demonstrated altering the patterning of the 3D printing process to create single-structure concrete with gradients in density, allowing structures with hard outer casings and light weight insides.

Concrete was a driving technology in making Rome– what will we get when we can 3D print with it at arbitrary sizes?

I haven’t had a chance to mess around with GDesign yet but judging from some of the things I’m seeing on the gallery page I really look forward to someone with a printer giving this machine a whirl.  The above example might be a little complex unless you’ve got one of those projector resin printers I keep seeing pop up here and there, but even knocking the divisions down one or two notches on this would probably generate something pretty awesome.

Okay admittedly there’d probably be some stray loops, but still, code up a building generator and rev up the automated build platform and you could spit out a little fictitious city overnight!

Not content with semi-consumable conveyor belts, Tom Martin found a supplier of titanium foil to produce a nigh-indestructable belt with excellent chemical, thermal, and mechanical properties!  On the belt’s page he writes:

Titanuim foil is an ideal platform material because:
1) Unlike aluminum foil it does not normally crinkle and maintains a smooth flat shiny surface unless deliberately creased. It it actually somewhat springy.
2) It is flexible like a sheet of paper and rolls smoothly through the mechanism
3) The coefficient of thermal expansion is much lower than plastic over the temperature ranges used, therefore although the plastic will adhere nicely when hot, when both are cooled the plastic shrinks more than the surface beneath it and pops right off.
4) It is physically extremely tough and will survive a lot of abuse
5) It is chemically inert and immune to solvents

Neat!