home - about - advertise     


Gehry, Dassault and IBM Too

By Martyn Day
September/October 2003
Reprinted with permission.

Frank Gehry is best known for the design of buildings such as the Guggenheim Museum in Bilbao, and the new Disney Concert Hall in Los Angeles. His use of advanced 3D CAD technology to help design and build these creations has been widely reported. By setting up an offshoot company called Gehry Technologies, it seems that this knowledge and capability will be available to other AEC firms.

Frank Gehry, the renowned international architect, plans to help the AEC industry re-architect its process with the help of Dassault Systemes.

According to the press release, Gehry Technologies (GT), has been created to ‘pursue technology and building industry initiatives.’ The aim of the company is to improve the way CAD tools are developed and used by the building industry, and to do this, the company states that it’s essential to foster changes in traditional building industry practice using these advances in computer technology. With the dynamic between partners and customers within the industry on the change, Gehry Technologies hopes to engage professional firms, product developers, owners and governmental and research institutions to share this vision of a more integrated industry.

Gehry Technologies has been created to transfer the technology that Gehry’s Practice has found to work on its projects, together with supplying training, and technical assistance for projects of other leading architectural, engineering and contracting firms. The company has research and development programs with academic and industry organizations including MIT's Media Laboratory, Georgia Tech, and CERF.

With the latest announcement that Gehry Technologies has signed up to re-sell Dassault Systemes’ CATIA - the system that Gehry’s own practice has used to great effect for many years - it’s hardly surprising that Gehry Technologies also has product development initiatives with IBM and Dassault Systems. CATIA is more commonly used in Aerospace and Automotive engineering than architecture and has very few AEC-specific features. However, we expect Dassault and Gehry to be developing AEC modules for CATIA over the coming year. As CATIA is a full-on 3D system, Gehry Technologies will be championing the move of the industry to 3D. The firm states that the current industry practice is still largely based on a two-dimensional, paper-based process, while the manufacturing (MCAD) industries have completely changed the way their products are designed, built and delivered by opting to use 3D at an early stage.

Over the last decade, Gehry Partners has adopted the use of model-centric design and has delivered a number of sophisticated, complex buildings worldwide. Armed with this knowledge and experience, Gehry Technologies sees a significant opportunity to bring the approaches it has developed to the rest of the building community.

Jim Glymph, CEO, Gehry Technologies commented, "This is an industry that does not have the cohesion of the manufacturing sector, and it must deal with processes, procedures, regulations, and even traditions of practice, that differ greatly from region to region, nationally and internationally. We are leveraging our international experience by identifying best practices and procedures that can use Gehry Technologies computing solutions to advantage."

Frank Gehry sees GT as an opportunity to extend his firm's contribution to the profession. "We've been working this way for years, and what we're doing is packaging our process and offering it for use by the industry as a whole. We'd like to see everyone have access to these kinds of tools, to open up new opportunities for designers and builders, and create a better environment for getting projects built.”

More questions

On the surface this all sounds great but when one thinks about the announcement it creates more questions than the available information answers. You have to wonder what the CAD vendors who develop AEC-specific modeling tools think of what is essentially an MCAD system being used as the high-end solution for the AEC market. Is Gehry’s process really applicable to 99.9% of other architects across the globe? And is this process, which appears to be based on Mechanical manufacturing transferable to the AEC community?

To answer these questions, I talked with Dennis Shelden, Chief Technology Officer (CTO) of Gehry Technologies on the company’s ‘raison d’etre’ - process and technology offerings.

Can you tell me a little about how Gehry Technologies came about and the company's mission?

Dennis Shelden: I think there's a good understanding of what Frank does and also a good understanding in the industry of what we have done with the available technology, even back when the technology side was hard and incredibly expensive. The key players in the practice, Frank Gehry and Jim Glyph, are really not desperate for business, they are looking at their position in the industry in an altruistic sense. Because of the demand for Frank's designs we have been able to rewrite the rules for projects and have a lot more influence in the projects. So these guys have a position in-terms of design and have relationships with academia and industry organizations and are champions of a new process, a new way of working. Frank's looking back on his career and would like to push not only for formal innovation but also process innovation.

There is considerable frustration in the US, about people wanting to do stuff they can't do - that's everyone in the process, general contractors, owners - everyone is frustrated. Frank and Jim are interested in looking for ways to promote a new way of working to get the industry into a better position. We have been able to get away with a lot of things, like changing the process. You have some sense in the UK the construction industry is probably a little bit more favorable to all the players but over here in the US it's a mess, so you can't do anything innovative!

We think the technology component of the process, could be a catalyst for trying different ways of working, in a controlled fashion. Over the last decade that we have been working with a variety of partners, we have been continually involved in setting people up to work with us, so we have acted like a quasi-technology company but that has never been formalized. We have been very frustrated with all the software solutions that are out there and think that they are all missing the mark and we are looking at a new set of issues; we are not just talking about generating and tracking 2D drawings. We have looked at everything that's out there and have gone with the most ambitious platform.

Gehry also does a lot of work with academia, the practice does a lot of research and there's a lot of research done on us, on how we do things. So we are trying to formalize a bunch of activities that we have participated in in the past, that are all fairly high-level functions.

So what do you think about the concentration on Building Information Modeling (BIM)?

DS: The whole BIM story is this year’s story, that's the new term. Without bashing everybody else's software, I would say that if there is a differentiation about our intentions, it's that we are really concerned with the Global process and while there are great tools about doing formal studies, walkthroughs etc. all these components are thought of as being the architect’s traditional activities. This has not translated to address the other 85-95% of the activities of getting a building built. It's not about producing another bit of software to produce walkthroughs to impress clients, it's about getting buildings built, which is where we have always been.

We are looking at what the fabricators are doing and what a lot of the general contractors are doing. The critical link is between the engineering team on the design side and the engineering team on the fabrication side. In a typical US contractual diagram of the process, there are 15 middle-men between those teams. So, can the software, without having the data pool becoming a complete free-for-all, allow for that necessary tight integration between design and fabrication?

When you look at that and compare that to the Boeing 777 process, that's kind of what you see. As Boeing is the single OEM, that controls hundreds of companies underneath it, there's obviously a huge difference to the building industry and that points to some of the issues that you want to achieve, people's roles will change, people's communication will change, integration should be stronger. We see different players push this model of design vs. building in a number of interesting ways, so engineers are starting to think talk about doing shop drawings, detailers are thinking about design services. It's all about recognizing that there is a tremendous risk and inefficiency in not getting good information upstream where you can get a better building. That's the battle that we are trying to fight.

But traditionally architects don't like taking on those high levels of risks?

DS: And there are good reasons for that, particularly in the US, where you can lose your company doing that! It's not about anybody's role in particular, as there are a lot of roles. It's a fictitious idea that a design team generates a specification for a building that is fully complete when it hits the general contractor. In the US that's where all the lawsuits happen is providing what is tacitly understood to be insufficient information and people discover that more information is needed, in a contract that dictates you provide a package that's sufficient.

There are a lot of issues that are about process and one of the places to point to is the contracting relationship and again that plays out differently in all the international arenas. One the strengths of Gehry's firm in tackling this is that we work in all corners of the world and we have a good insight as to how things play out differently.

But you aren't a normal AEC firm, you drive the process?

DS: It would be crazy to compare us to someone like Bechtel that are fully service, design/build organizations as the model is vastly different. It's one thing to say we can call the shots on our projects; it's another thing, but perhaps a natural progression, to demonstrate to the market that there is a different way of doing things.

Part of the charter for us is to develop metrics that substantiate that there are better ways of doing things. The result will be better projects and vastly more economical process, in that what you are doing is reducing risk, so risk has a price and if you can control risk and people in the field can snap parts together and they fit, then you don't have to stop construction while someone cuts a whole in an I-beam. Owners get caught because they pay for these errors.

By creating a complete Virtual 3D model you eliminate the risk for errors?

DS: The notion of a comprehensive model is not exactly the situation, there's this whole notion that 2D is dead, long live 3D, but that's not really the case either. Parts of the success of this firm has been about tacking geometrically complex buildings and using fabricators who may want full size plots for template cutting - and we can provide that. If you have a structural model of ribs that are floating in space and ask people to price it, you are going to get a big price. But if you can use software to process the information and produce information that they are familiar with, then the price drops. It's about the ability to process and transform information.

Having said that one of the biggest benefits is being able to do an integrated, coordinated model. We don't produce all that information, we bring it together. Just the fact that the design team can review shop drawings in an integrated 3D world with a lot of attribute information on it, as opposed to a stack, a foot high, of engineering drawings, is a clear win and that's the obvious benefit. You can take everyone's shop drawings and engineering drawings and position them somehow in 3D - be they 2D or 3D objects or structural elements and be able to do clash detection and interference checking and just visualize.

If an I-beam skims through a surface there are tools to find those instances. When we talk about visualization those are the things that are interesting to us, not including plants, trees or little dogs! It's all about the construction information.

What proportion of the architectural market do you think Gehry Technologies’ message will appeal to?

DS: Good question but I don't think I'd throw a percentage out there. If you look at the US construction market and look at the bottom tier of wood frame housing and assess the inefficiencies within that. There's not a lot of inefficiency, but you pay the price of very constrained design. So, how big an issue is that for the developer? It depends on the developer. Clearly the benefit is at the level where things are complex, the sort of building we do but there are plenty more complex projects.

In Frank Lloyd Wright's time, all you needed was nine drawings, and now you need a thousand drawings for the same projects. So, clearly the demand for performance and codes and requirements has outstripped that technology and it shouldn't be surprising for anybody that there's trouble with complexity.

But these codes have been put in traditional AEC packages. Your chose of system, CATIA doesn't have any as it's not tailored to AEC?

DS: Dassault is one of our strong partners in this effort. They do for certain systems have perfectly viable systems for structural and mechanical. You could argue about the positioning of them, but they are there. In terms of the conventional architectural information, it's probably a subtler problem and then you have to look at what the architect is designing. I think CATIA is acknowledged as having, if not the strongest, then one of the strongest geometric and knowledge-ware parametric engines. When you look at the amount of information created with buildings, they are complex and the fact that people may not perceive that complexity because the models aren't quite there yet, doesn't mean that that the information is not floating around.

In terms of the differentiation, even in terms of geometry and the way things sit together, the way the intelligence propagates from an architect’s view, to an engineer’s view to a fabricator’s view, a lot of that hasn't been worked out in the industry and it really is a very complex geometric problem that I think has not yet been solved.

When you look at our buildings which contain 10 to 50,000 structural plates and I-beams with corresponding numbers on the curtain wall and ductwork and all this information floating out there, the capability to integrate all that information is just not there in the industry yet in the conventional AEC packages.

But to use systems like CATIA you need programmers? Autodesk’s, Bentley’s and Graphisoft's solutions all hide the complexity of modeling?

DS: I think increasingly that you need both engineering knowledge and computing knowledge. There's been a tremendous increase in the availability of really smart computational people all over the place. I don't believe the goal is to develop one product that beats everything else. The goal is to build an integration environment that can host and control and integrate all the information.

I don't want to sell short the fact that we need some sort of architect's knowledge to enable them to do the work. The question is, can one develop an environment where all this information can sliced, diced and transformed.

With so much work to do, how much development will Dassault Systemes, creator of CATIA, be doing?

DS: I have to say that we are not in a position to talk about that plan and any sort of timetables, but there is a plan. We have a long-term plan and short-term deliverables. We have already done a lot of work and we also have tremendous resources in our notion of the process. Most of our projects have been fully integrated, through design, engineer and fabrication, not exclusively on CATIA, so there is a lot of focus on broad-based integration and translation standards.

Surely the cost of doing all this for architectural firms is hard to justify?

DS: I am not sure the architects will be the people that single handedly change the construction industry. It needs a lot of players to come together to recognize that things have to work differently and I think the owners are fundamental to that.

And one of our challenges in the near term is to demonstrate - particularly to the owners - that the way things are working now are costing them money and time and ultimately they are the ones that get the benefit. The question of what will 'incentivize' architects I think is going to be treated on a regional basis. And the fundamental way into this industry - and this is core to the Gehry approach - is to talk about Projects, not about architects or any particular player in the mix. We have to be able to convince the whole team to think about another way to do that, and I think it's the big challenge, it’s the contractual issue.

CATIA is still expensive when compared the offerings from traditional AEC CAD vendors?

DS: We are hoping to work out something on the price, there's going to be some need for reconfiguration to make it work on this process. We recognize that the players in building industry are different to Dassault's typical engineering customers.

I talked to Carl Bass, Executive VP at Autodesk, he said using CATIA in architecture was like using a jackhammer for orthodontistry!

DS: Well, I guess I wouldn't expect him to jump on the wagon. The jackhammer paradigm…. I would take exception that architecture and the AEC industry is such a dumb profession that good tools won't have an impact. We think differently; the financial and incentive justification needs to be there, and I think it is. The industry dynamics have to change in order for it to work but just look at the huge amount of money that's being lost of construction projects today, tens of millions of dollars. So the money is there and if you can crack this pervasive inefficiency and there are all sort of metrics get quoted to highlight this - over the 90s the building construction to architecture market was the only segment to make a net loss in productivity. It comes back to the vision of Frank and Jim and we are going to shoot for it.

About the Author

Martyn Day is group editor of MCAD Magazine and AEC Magazine.

CADdigest Weekly
Email (required) *

A week's worth of articles will be emailed to you. See latest issue .