The Futureof CAD

I spoke in front of the Vancouver AutoCAD user group last night on the subject of "The Future of CAD." One woman came up to me afterwards to complain, "I thought you were going to talk about which buttons to push. That's the only reason I came."

Some of the things I did talk about:

To better understand what might happen in the future, it is important to know the past. This helps us decode what is happening, and what is likely to happen next, based on common human reactions. For example, if a software vendor switches from selling to the masses to selling vertically or repackaging the product as an API, we know the product is probably failing. Same if the price goes from $199 to $99 to $49 to free.

Computer hardware has created limits to growth. CPU speeds have stalled at around 3GHz; programmers are not taking advantage of 2- and 4-core CPUs. RAM is maxed out at 2GB/3GB. Hard drives get larger but not faster. Flash drives are too expensive, and when they fail, fail 100% with no data recovery possible. 64-bit computing requires 64-bit CPU, 64-bit OS, AND 64-bit apps -- and then all you get is the ability to have more of the CAD data in RAM.

There is little innovation going on in the CAD world. A few new tactics crop up here and there.

The Internet made CAD vendors rethink themselves, because the Internet allowed the switch from drawings on paper to drawings as data (DWF, PDF, et al). This lead to the concept that CAD is not about drawings, but about data, which lead to CAD vendors thinking of themselves as data enablers. Hence, companies like Dassault and PTC no longer using the term "CAD," but other convoluted phrases.

The CAD market is plateauing, and so CAD vendors are looking for more ways to sell more software in all four directions: earlier in the design process (conceptual and early marketing), afterwards (PLM), down to the shop floor (CAM, etc) and up into management (viewers).

In terms of users, bifurcation is occuring: (1) those who continue with tradCAD, like AutoCAD, currently 70% of the market; and (2) those who split off to 3D modelers, currently 30% of the market. Perhaps this generation has to die off before tradCAD disappears.

Some reaction from the audience:

- the limits of computer hardware don't affect them.
- perhaps when 30-inch screens become cheap enough, then D-size paper will no longer be needed.
- for most of their work, AutoCAD-style CAD is just fine.
- CAD needs to make greater use of Internet-generated technology, such as using Second Life to bring distant clients together on simulated 3D projects designed in CAD and placed into landscapes.

(Thank you to Dave Witso for inviting me, and to Bill Fane for driving me to Burnaby's BCIT and back.)

Comments

[matt]

Ok, I'll take the last bullet point about Second Life as sarcasm, I hope. At the worst it represents a fear of the future.

If you aren't seeing innovation in the CAD market, you're under a rock or in bed with the wrong product. I keep hearing about how "mature" the CAD market is, but everyone I know using a certain 3D modeler keeps begging the developers to slow down because we can't keep up with the new functionality! I'm talking CAD functionality, design, visualization, shape creation, things like that, not useless web-connected bs.

Granted, AutoCAD is old news, and if you're so deeply involved with that, there can't be much innovative happening, especially when you expect it all to happen on D size paper.

I enjoy most of your posts, but fear that this one has gone square off the rails.

Hoping you recover...

[Devon Sowell]

Hi Ralph-

Interesting comments, but I disagree with most of them.

RE:"for most of their work, AutoCAD-style CAD is just fine." Not for any of the 24 companies I Consult for.

Devon
http://www.3-ddesignsolutions.com

[Henrik Vallgren]

"programmers are not taking advantage of 2- and 4-core CPUs"

That will change ... it's a promise. While traditional multithreading settle for a fix number of tasks, the new challenge is to divide computing onto an unkown number of cores, and we're still learning here. As in many other disciplines, we've got too many options to choose from.

[Jon Banquer]

Ralph,

If better documentation was made available so you could try doing some very advanced surface modeling of say a car body with a product like T-Splines for Rhino, I honestly don't think you would feel that there is very little innovation going on in the cadcam world at present!

To see how innovative T-Splines for Rhino really is check out this video:

http://www.youtube.com/watch?v=nZYFqjnF-_I

Here is an example of what I mean when I say much better documentation for cadcam is badly needed:

http://tsplines.com/resources/car1.php

What is the point of showing someone how to do a tutorial without the starting 2D prints or pictures? To me this has got to be the height of stupidity and by no means do I mean to single out the T-Spines people for this. I'm just using this as an example.

I just got SolidWorks product manager Mark Biasotti to release the pictures and sample files for the really nice advanced surfacing video tutorial he did on using SolidWorks for consumer product design. (Thank you, Mark!) Mark’s advanced surfacing tutorial is the only advanced surfacing tutorial that I know of that is even available on the Web for SolidWorks! Just how ridiculous is that? This is typical of the condition that currently exists in regards to poor documentation availability.

I’m of the very strong opinion that the only way someone is going to be convinced of how good or innovative a cadcam product is if they can actually do what they see being done themselves with step by step hands on tutorials.

Jon Banquer
San Diego, CA

[Rob Rodriguez]

"To better understand what might happen in the future, it is important to know the past."

This is precisely the problem with the future of CAD. We have too many people living in the AutoCAD or TradCAD past. These people are either unwilling to step into the future or don't realize AutoCAD or TradCAD has no future. It has reached it's peak.

[R. Paul Waddington]

"To better understand what might happen in the future, it is important to know the past." Spot on!
Rob said, “This is precisely the problem with the future of CAD. We have too many people living in the AutoCAD or TradCAD past. These people are either unwilling to step into the future or don't realize AutoCAD or TradCAD has no future. It has reached it's peak.”
Really Rob? The problem with the future of CAD is to a huge degree the fault of those who have elevated 3D CAD above its station as design and draughting tools. CAD, in any form, without a designer or drafty using it is absolutely useless; and CAD does not make a good designer or drafty and it never will!

The relentless push by those who want others to “move on to 3D” is wasted effort. If what you believe is true it will happen because of what you do not what you say. 3D will demonstrate why it may be worth moving forward to some and to others it will never be profitable and therefore they should not use it.

As I have said on other occasions 3D CAD has been around for as long as 2D CAD so it may pay those who “cannot understand” why some stick with 2D to have a long look at what they do, and learn something in the process. Making money is why we design, draft and manufacture and if 2D remains doing that for some then they would be fools to follow the 3D pack.

Also there seems to be an assumption that those that use 2D have not looked at or cannot understand 3D; it will come as a real surprise to those with their heads buried in the sands of 3D to find out that that is not always true. For those that have looked and know exactly why they continue to use 2D, predominantly, are you suggesting they should forgo their profitability simply to please 3D zealots?

30% market share after 25+ years may well mean that 3D does not have the profitability component, for all, that it may have for those it suits and who use it; designers and drafty cannot be all fools and I don’t think it serves the CAD market place any favors to continue telling 2D users they are fools if they don’t use 3D.

I have and use current AutoCAD and Inventor licences and AutoCAD continues to earn, as a tool, more than Inventor; Rob when would you suggest I reduce my profitability to placate the 3D mob and market or would you just prefer I turn off my life support system because after after 42 years in the design world, 35 of those closely involved and using CAD, but because I still use 2D I don’t know what I am doing, am irrelevant, and therefore should just go away so only those who use 3D can get on with their 3D modeling life?

R. Paul Waddington.

[Paul Marsh]

I think that commenters should double check Ralph's post and note two important facts;
1. The title is, "The Future of CAD" and
2. the audience is an AutoCAD user group.

The future of CAD to a general CAD audience would certainly have a different flavour, or at least I would hope it would. In the same way that a talk about technical writing for a web audience would be different from a presentation on the future of technical writing for an MS Word user group. There is an implicit bias to the talk.

I think that MS Word and AutoCAD have a lot in common; wide spread market adoption, little innovation and a radically changing marketplace, (ie. innovation is happening elsewhere).

>- for most of their work, AutoCAD-style CAD is just fine.

Compared to what?

For a while, messenger boys suited people just fine - manual switchboard was slower then a 12yr old boy with a paper note.

[Matt Stachoni]

Sorry, Ralph, I have to disagree with your idea that "there is little innovation in the CAD world." Indeed, I think we are drowning in innovation. As we move to vertical applications, most of us are in a constant state of catch-up, and is a primary mover that results in events like Autodesk University growing by leaps and bounds every year.

And sub-12-month product cycles aren't helping.

Even in plain old AutoCAD, new and "disruptive" technologies are introduced with each release. Of course, whether you consider something like Sheet Sets, Fields or scalable annotation to be "innovation," or simply a case of "well it's about time" is debatable. What is not is that they change how we do things by leaps and bounds.

I certainly do not view the CAD market as "plateauing;" if anything, I think the advances were are just starting to see will continue to develop exponentially and the toolsets to build more effective CAD applications are built - in AutoCAD, yes, but particularly in the vertical markets.

For most folks, the move to vertical, specialized products is inevitable, as the industry-specific tools are developed to meet those specific needs. The BIM movement is a prime example of the AEC industry driving innovative software development (as piecemeal as it may seem). This had already happened in the mechanical design and manufacturing fields. Similarly, any push to 3D is essentially meaningless without industry-specific context, which is why Revit, ACA/ADT, Civil3D and Inventor are all so popular. If anything, innovative advances in vertical apps should migrate down to horizontal catch-all apps like AutoCAD where appropriate.

Likewise, your notion that a product is in deep trouble if it goes from $199 to $99 to free is flawed. Look at web browsers. Look at high-end open source projects like Apache , Java, the GiMP and Blender. The quality of free or cheap software is amazing, where the quality of some very expensive applications is, well...not so much.

Take SketchUp, which went from $500 to free overnight, where the zero price lowered the cost of entry and expanded the market itself exponentially. Now you have thousands of people using it who don't do CAD or 3D as part of their daily job, but use it to solve all sorts of problems. When you have grandparents who can barely turn a machine on without help designing their kitchens with 3D software, THAT is something truly special. Remember our conversation about how everyone and their dog was going to model their house in SketchUp and put it in Google Earth?

And it is not hardware that is limiting growth, it is software. Currently the state of hardware is off the charts relative to what the software is doing to feed it (with the possible exception of idiot Vista, which will kill the fastest hardware just playing an .mp3). Many of us now have dual-core, 64-bit CPUs but are still hobbled by 32-bit operating systems - mostly due to driver development and 32-bit software compatibility - and single-threaded applications.

The problem is twofold: A chicken-egg syndrome where developers won't create 64-bit software if there is no perceived demand, and customers who cannot buy 64-bit software (and thus create no demand) because there's nothing available to buy.

Second, multi-threaded applications are really hard to build efficiently and ensure a stable product - and stability trumps performance when upgrades are at stake. And with many applications - AutoCAD included - multithreading across the board may not result in a whole lot of performance benefit (it's already multithreaded where it counts - in rendering and display).

What we do have is program after program installing useless resource-sucking services and startup programs (I think the technical term is "craplets") that need squeegeeing off your system before it operates like it should.

Finally, the Internet's effect on CAD "innovation" is approximately zilch (sorry, AutoCAD 2002 developers!). With the possible exception of allowing software companies to leverage the "always connected" status of their users to perform intrusive license checking.

But folks exchange raw CAD files when the exchange of real CAD data is really important. PDFs and DWFs are simply nice ways of safely and efficiently packaging sets of sheets in a manner that doesn't allow editing by the recipient, or requiring a $4,000 piece of software just to see something. But neither PDF or DWF need the Internet per se, except again as a transfer medium (and most often, through lo-fi inefficient email).

It is more accurate to state that "The Internet made CAD vendors' MARKETING DEPARTMENTS rethink themselves," because for several years you had everyone rushing to integrate the Internet into their product lines because it was hip, without any real purpose or overriding idea of exactly why.

[Kevin H. Driscoll]

Your readers will be amazed at their ability to access their Cad files on their PC from any iPhone. Clearly as the movement toward remote access and mobile computing moves forward, this represents a substantial value in terms of 24x7 access.

[Paul]

Matt, some engineering projects (in fact, many of the largest ones on the planet) are still built by construction guys/girls working in the field. They still work from D-size prints.

Computer-controlled robotic welding and erection of piping and structural steel just isn't practical.

I don't know what kind of CAD or modeling you do but I get the impression that you're only considering your own little corner of the engineering world.