Ricardo Bittencourt

Are you familiar with CAVEs? I worked on one of them back in the early 2000s:


Seems like CAVE is a bit space intensive, and somewhat limited (I can't see managing to do a floor and ceiling at the same time, unless you don't mind putting at least one projector in the cave with you, or make the floor of it an enormous and very durable screen). It's certainly doable, but it seems exceedingly expensive. As in, hard to fit in the budget of most engineering consulting firms.

The Oculus (a mass-market VR Goggle rig recently funded on Kickstarter) or a similar system seems like a more cost-effective solution. Of course, that requires an interpreter program to get from a CAD 3D model/render to something the goggles can display. I don't think it's available yet. That said, the Oculus has gotten a fair bit of attention of late, so it's quite possible someone will rig something up soon. After all, most 3d compatible CAD programs already do the tough part - making and rendering a 3d model. And it's probably possible to get a walkaround from the 3d model using the Source engine (I'm not a modder, so I don't know how easy that would be, but it seems logical that it's possible to feed Source (or Unreal engine or CryEngine or whatever you want - I'm only saying Source because it has some nice modding tools available) a prerendered 3d model and make it navigable in some form. From there you just need the drivers for the VR goggle compatibility, which Oculus is supposed to deliver.

On-the-fly rendering, however, takes a bit more work. If your client is fine with interpreting wireframes, it's doable, but if you need to supply renders (and let's face it, unless you work with wireframes on a very regular basis, it can be hard to interpret, so most clients would need renders for a virtual walkaround model), you'll need a really nice computer to do on-the-fly editing (since rendering takes a long time). (Additionally, my modern tech setup of AutoCAD->Source->Oculus would need some adjusting since, even with live rendering, you'd still need someone outside the model making the changes to the CAD design and sending a new 3d model to the rendering engine.)

This kind of system is probably more useful for architect and civil engineering scales than mechanical (for the most part) since most people are not yet sufficiently used to the idea of using VR to expand a hand-sized component to something you can walk into, but that's largely just a matter of training the mind. (I'm sure that if you showed a 8.5x11 paper printout of an ant in scale to most medieval citizens (aside from a few of the more talented scholars of the day), they'd have trouble grasping the concept, whereas today most people can grasp that it's just in scale. (You may also have issues regarding time travel paradoxes, language barriers, and social conflict of myriad forms, but that's beyond the scope of my example.))

Wow...did I really just type all that?

motan's picture

CAD approach is nice, but it's still not a physical representation. I was thinking something along the line of holograms (like here: The holographic maps that I've seen are pretty awesome and great for planning, although you're not walking inside unless you pay massive amounts of money for a big one.


The first comment's CAVE method can do that too, and it's pretty old technology. You just need a big room, some time to build the rig, a number of projectors 3d glasses and some specialized software. Unfortunately, that still adds up to quite a bill. The VR goggle option I mentioned simply requires a set of goggles, which are supposed to be mass marketed soon, a compatible game engine (apparently the conversion has to happen in the game engine and not, as I was assuming, in the device drivers). The idea is that you see and hear the virtual environment and not the real one. You may not be able to stand up and walk through it like the Matrix's VR system, but it's probably our current best bet.

Also, every hologram technology I've seen has been employing some kind of trick - a rapidly spinning mirror with a very high refresh rate, for instance. Walking alongside or within one is, I think, still in the realm of science fiction. That said, I suppose you could build a room with VR rigs on all sides, thus presenting the feeling that you're in a large empty space embedded within a single very large projected object. As you said, that would require enormous sums of money.


I smell a business opportunity. Rent such a space out to engineering consulting firms?


Sounds like a good idea. Unfortunately, there's nothing proprietary about it (or rather, the patents are held by other people) and thus it's somewhat difficult to attract investors to build up the necessary capital to start. It would also require a rather focused marketing campaign and a fair bit of market research to choose the ideal city to start in - obviously small towns are out, but where does one find the engineers most interested in this sort of service.

Once all that's done, you need to figure out how to make a margin - you're paying rent or taxes/mortgage on your site, you may be paying down a loan in order to finance the equipment needed, and you need to fund your marketing campaign (which can be a lot of money), but if you make the price of the service too high, engineering firms still won't be able to pay the bill.

If one were to actually do this, the CAVE is probably the most effective current model. Holograms look really nice, but are horrendously expensive. Back-projected screens, a projector array, and some manner of fancy 3d glasses (the fundamental components of a CAVE) are much more affordable. If you wanted live editing, you'd also need to develop some means of integrating CAD editing tools into the 3D environment (expanding on the Wii/Kinect technology, maybe, or look into what Sixth Sense does.) Depending on what the users want, the rendering could also be an issue. (Generating a plane is easy. Applying a premade texture is easy. Rendering it fully with realistic light sources and shadows and so on is not. That said, if we're renting a large space in a major city and buying the equipment necessary to run a CAVE system, a computer with lots of extra processing power should easily be affordable. (That said, the top processor out there at the moment appears to be Intel's top line E7 server processor, at somewhere over $4000, so perhaps not))

Too big a project for me to be a primary in (I'm more interested in technical problems than business projects), but if you like it, go for it.


Heh. No, I'd be terrible at business. However, I'm sure some businessperson or some successful guy would enjoy the challenge.

chuck's picture

We used CAVE in my engineering firm as early as 1998. VR blackout googles and a hot "glove" to allow interaction with the VR enviroment. It was fun to watch a client strap in and step over imaginary cable trays, piping chases, etc. We used it primarily for process engineering - to maximise productivity from a fixed asset. Our clients were some of the world's largest corporations (at the time - times have changed) and VR technology allowed us to hold onto our existing client base.

We had a coroprate jet, too. Both CAVE and the corporate jet have gone out of vogue in the engineering industry, although each had their place. Don't overthink the app - rather strip it down to its basic element - allowing those unable to see 3D images from 2D representation - and figure out an inexpensive way to make this happen. Positioning software and accelerometers in a dark box will get you 80% of the way home without having to worry about shadows and the fragrance of night blooming jasmine.

Thnikkaman's picture

I actually develop a desktop application that takes 3d CAD models and allows you to take them apart and put them together in real time. We also have a low cost 1 wall cave and motion tracked HMD that allows a user to interact with a design. I believe that we put everything together for under $20K.

You can also check out this:

Matoyak's picture

Maybe look into the Oculus Rift. Yes, it is intended for Gaming, but I can't see why it couldn't be applied to CAD based design

waldosan's picture

you should think about designing the system yourself, if this is such a common desire for CAD users and Engineers alike then i'm sure there would be a sizeable market to actually make this work. and it's not like you are short of ideas, every engineer worth his salt has a method for how he would build tony stark's holographic table thing.


Actually, not so. Some of us work with things where a holotable isn't a very practical design tool. Mechanical and Electrical Engineers often work at a scale where such things are useful (though not always - a hydro dam's inner workings are electromechanical and larger than the average house). Civil, Geotechnical, and Mining Engineers all work with enormous objects all the time. And many of those in the latter two categories (and some in Civil) work pretty much exclusively in tunnels surrounded by large amounts of rock, the exact properties of which vary significantly and are critical in overall design. Scale 3D models of underground mines viewed from without are not sufficiently useful to call for such a device. A VR walkaround system of the interior of tunnels (as per the Oculus Rift ideas) would be far more useful.

That said, unlike me, Angela _is_ a Mechanical Engineer, so she might have some ideas. Though actually building one is primarily a software problem (augmented reality goggles of some form aren't all that mammoth a design problem - I don't see how they could be that much harder than the HUD system used in combat aircraft), but the software needed to render the graphics, interpret gestures properly, and tie that collection of hand gestures into a fully functional CAD toolset is. As for projecting a full hologram that you can stick your arm into to virtually try on your gauntlet of doom, I don't think even a physicist could give you a method to start from in which it's physically possible, which rules out most engineers from working on the problem. Scientists come up with the new science, Engineers come up with how to make that into usable technology. (And then there's the Engineering Physics people, who blur those lines, with interesting results.)

And finally, taking the long road and independently developing something is a common mistake. If you want to know how long the full length of cable in the Brooklyn Bridge is, you should look it up before you fly to New York and break out the surveying equipment (this example chosen for it's absurdity). Similarly, if you want a VR/AR model of something, you should first look up what VR/AR technology is out there, and only develop a new one if what you're thinking of can't be done with pre-existing technology.

SuperSparkplug's picture

You haven't happened to have heard of CAVE technology, have you? For the most part, it allows you to do just that. Sure, it requires LOTS of money, space, and programming skills so is not feesible to make, but there are quite a few academic and research places, such as Universities, that have them hidden somewhere.

I actually work for a research lab at Ryerson University that has a CAVE. It's pretty wicked! :)

CrazyAlmostCanuk's picture

Oddly enough, there is a game accessory manufacturer out there (DTX, I think) who makes a visual headset specifically designed to interpret 3D/2D games into full depth 3D (if I wasn't clear, I mean it's designed to take games designed to be displayed on a 2D screen, which make use of 3D modelling, and actually add the depth only approximated on the standard screen)

Since these are *already* compatible with any computer graphics card that supports 3D imaging, and with the PS3, I can see where there would be some level of difficulty adapting most current CAD or 3D modelling software appropriate for engineering displays to them, but would it not mostly be a problem of calculating the offset of each "screen" to provide true depth from wireframe images?

If this is the case, then clients and co-workers not only wouldn't need a huge presentation room like CAVE, but could teleconference from "inside" the model, with the hardware running under $500 per unit, commercial.

But maybe I'm over-simplifying. I haven't even touched a piece of code since 3 dimensional monitors and TVs became available.

Vanessa Barratt

Might sound stupid but why not build it in secondlife. It's not the same as being in it but the client would have the ability to move round something and see it from all angles without the need to travel, no special equipment and after very little tuition on the camera and movement controls.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

Thanks to our Superstar Patrons!