Evan on Engineering Software

Don't play the video below before you've read the text.

Deke Smith is the executive director of the buildingSMART Alliance, an  outgrowth of the  NIBS (National Institute of Building Sciences) National Building Information Modeling Standard initiative.

Yesterday, on the buildingSMART listserv,  Judd Peterson (an architect now working in the facilities management software business) pointed out an article from Engineering News Record: Lack of Interoperability Is Biggest of Many Gripes

Deke wrote back today, saying:

I did find this interesting and actually, I think it should be very helpful to our cause. It clearly identifies that we have a problem in the industry.  In fact, we believe it to be a $300B problem.

We also have a plan on how to fix it! Hopefully this will help gather the resources we need to change things. We have done VERY well to this point with virtually no resources. I would hope that no one really expects us solve a $300B problem with $300k!

Certainly an incredible ROI, but not very realistic...

Stop and think for a moment:  Interoperability, just in the building sector, is estimated to be a $300 Billion problem?  That's such an astronomical number that it's not even worth arguing about.  Who cares if it's high by even $100 billion dollars?  It's the kind of problem that demands attention.

Over time, I've thought about this problem a lot... And I think I know who could actually solve it.

Consider this potential scenario:

It's April 2009, at the Congress On the Future of Engineering Software, in Scottsdale Arizona.  At an executive conference room, away from the crowd, a handful of powerful industry executives are gathering.  Carl Bass, from Autodesk.  Bernard Charles, from Dassault.  Greg Bentley, from Bentley Systems.  Tony Affuso, from Siemens PLM.  Dick Harrison, from PTC.  Halsey Wise, from Intergraph.  Georg Nemetschek, from Nemetschek AG.  Richard Longdon, from Aveva. Shantanu Narayen, from Adobe. Sam Palmisano, from IBM.  Larry Ellison, from Oracle.  Henning Kagermann, from SAP.  Steve Ballmer, from Microsoft.

As each of the executives tries to find a seat where they're not sitting next to someone they don't like, Al Gore wanders in. 

All of a sudden, there's silence, as everyone realizes who it is.  Gore smiles, and says "Don't mind me.  I just came because I was curious who managed to convince you all to sit down in the same room together.  Maybe I could get them to help me with the global warming issue."

After a few chuckles, everyone manages to find a seat.  It's strangely quiet, considering the egos in the room.

Almost as if on cue, Dr. Joel Orr, the gentleman visionary of the CAD industry, walks in, pulling the door closed behind him.  Looking relaxed, and wearing a festive shirt, he smiles genuinely, and starts to addresses the group.

"Ever since the second computer program was written, there has been an interoperability problem.  Great minds have tried to solve the problem, but to no avail," Orr says.  "The only true solution is for software vendors to openly document the structure, syntax, and semantics of the APIs and file formats used by their products.  Yet, despite the hundreds of billions of dollars in costs to your customers every year, and despite some of you facing FTC and EC investigations and hundreds of millions of dollars in fines..."

Orr is suddently interrupted.  From the back of the room. Ellison shouts out "Hey, Ballmer, better you than me!"

"Play nice, Larry," cautions Joel, ever the diplomat. "Everyone here has some blame in this."

Orr continues. "It took tremendous work to get all of you here today.  But, it will be worth it.  We have found the one person we believe is capable of convincing each of you that it is in your best interests to fully embrace open interoperability. This person will be facilitating this meeting."

There is confusion in the room, with everyone looking around, trying to figure out who the facilitator is.  Gore sits in the back of the room, trying to keep a straight face, as if he knows something no one else does.

Walking towards the door, Orr continues speaking. "I'll be leaving now, so you can get started.  I'm quite confident that, within the hour, you will have a definitive agreement on interoperability, and the problem will be solved, for good."

Pausing just for a moment, and taking in the incredulous faces around the room, Orr opens the door, and says "I take great pleasure in introducing your facilitator...  Please welcome the man who will solve the interoperability problem...

(Play the video now.)

If you don't read the blogs at http://cofes.com/Community/Blogs.aspx, you're missing some thought provoking essays, from people who actually know what they're talking about.  The following is from Brian Seitz.

What is called for is a breakthrough in the product design process

In the 70s and 80s, Engineering and Manufacturing Software developed upon the technological basis of geometry capture. This is essentially the digital equivalent of using the Cartesian coordinate system and descriptive geometry that architects and engineers in previous decades had been using only now created in cyberspace. Further refinement to this approach yielded such benefits as 3D Wireframe, Surface, Solids and Parametrics and to some broad sense of the word Features. Additionally the geometric definition approach along with advances in computer technology and computational mathematics to product design yielded other advances such as analysis software applications in (Structural, Kinematics, Human Factors, and Fluid Flow)

What this approach and incremental refinement has not accomplished over the years has been a breakthrough in product design productivity. For all its improvements, the product design process has improved little from the methods developed more than a century ago.

Evidence of this lack of progress is the continued inflexibility of product designs, increased levels of efforts to integrate components or products –if possible without significant modifications—into larger systems or solutions. Design System sensitivity to requirements or performance changes is still a major issue. A simple requirements change can create significant redesign, rework and scrape of products. This fact has not changed since its effects where documented in a US Air Force study OVER THREE DECADES AGO in which it was found that more than ninety percent of a products costs and inflexibilities were created before the initial geometry was completed early in the design process. This infers that CAD systems are not really product design tools but rather product description capture and refinement for manufacturing produciblility purposes [A discussion of another time].

The ability of engineering firms to envision potential changes forecast impacts and adjust to these changes is a significant weak point for firms as they enter this age of rapid change. Attempts to “plan for change” and technology flexibility have not yielded the hope for benefits either.

What is truly being called for is a breakthrough, a fundamental change in how products are design and produced. The stepwise refinement of today’s design systems will no longer yield the productivity corporations are demanding of the function.

Did you ever think you knew something, then realized that you didn't?

I keep on seeing people talking about "DWG", but they never seem to say what they're talking about.  While I think I know what DWG is, I've come to realize that what I think it is may not be what other people think it is. Or isn't.

Rather than starting off saying what I think DWG is, I'd like to ask you, as an open question: what do you think DWG is?

Oh, and if you think this is a trivial and meaningless question, you should know that attorneys are already staking out positions on its answer. 

BBC News reports: 

"The growing problem of accessing old digital file formats is a "ticking time bomb", the chief executive of the UK National Archives has warned.

Natalie Ceeney said society faced the possibility of "losing years of critical knowledge" because modern PCs could not always open old file formats."

Look at the CAD industry?  Which CAD programs are creating the most significant time bombs?  CATIA?  Pro/E?  AutoCAD?  SolidWorks?

I've raised the alarm bell on this issue for years... yet, neither vendors nor users seem to grasp its importance. 

The term "innovation" seems to be increasingly in vogue these days.  For example, Amy Rowell, a friend from the days when I first started writing about CAD, is executive editor of Innovate Forum, an on-line magazine about innovation.

Over the past year, when I've heard people talk about innovation, I've sometimes felt a bit like the child in the Hans Christian Andersen tale, the Emperor's New Clothes.  I find myself asking the seemingly simplistic question "what do they mean by innovation?"

I was reading a paper tonight that helped me to understand this rather basic concept a bit better.  The Structure of Invention, by W. Brian Arthur, has a rather obvious theme, given its title, and starts off with a nice clear statement:

Schumpeter famously divided technological change into three phases: invention (the creation
of new technologies); innovation (the commercial introduction of new technologies); and diffusion (the spreading of new technologies).

In this case, Schumpeter is the Austrian economist Joseph Schumpeter,  a contemporary of Keynes, and regarded by many (including, for example, Peter Drucker) as a modern prophet.

Schumpeter's definition of innovation is quite a bit more involved than Arthur hints, but has as a central feature the commercialization of new technologies.  Innovation is a much more subtle thing than invention.

In following a few leads on Schumpeter, I found that there are a wide variety of classic definitions for innovation, and just as many variant types of innovation.

One interesting definition from the realm of network theory, as posited by Regis Cabral:

"Innovation is a new element introduced in the network which changes, even if momentarily, the costs of transactions between at least two actors, elements or nodes, in the network."

If you look at the world of CAD as a network, connected by shared data in common formats, this definition takes on an interesting significance.

Wikipedia has a rather good article on innovation.  Definitely worth a read, if you find yourself thinking about these things at 2:30AM on a holiday weekend, as I occasionally do.  While you're there, check out Warnock's Dilemma, which will nicely explain why posts such as this get almost no comments.