Gabriela Celani: What is designtoproduction?

Arnold Walz: Designtoproduction, first of all, is an architecture firm, the only thing is that we don´t do design. Our clients are architects who want to build complex geometrical structures. They came to the conclusion that they don´t need to have an in-house special design group. For example, Renzo Piano sometimes builds freeform, but mostly he builds less complicated designs, so he doesn´t need a special geometric group and he decided to work together with designtoproduction.

What we do is we help realize these complex structures in two ways. The first way is supporting the design phase; we write scripts which create the forms. The scripts are very flexible so we can try out different façade forms, beams, positioning systems, whatever we want to try out in order to optimize the overall design. The next phase is when the design seems to be ready and the team agrees that it is it, but you come to a stage when you have to incorporate all the aspects of structural engineering, building physical aspects. These things also cause changes, and these changes have to be handled. In these complex buildings a small change can have dramatic changes all over the place. In the Mercedes museum we once changed the central core and it even affected the façade. From the view point of conventional planning, these changes would have been a catastrophe. For us it´s not a big deal because it is all laid out in a way that you can update everything by just changing a parameter.

Maycon Sedrez: What are the files that the architects send to designtoproduction? What are the documents that you receive? Which are the analyses that you do in those files?

It varies a lot. For example, for that façade in Davos, we got a clay model that showed the design intention. Normally we get 3D files, in Rhino (1) most of the time. But usually the quality is not good enough to base our programming on top of that data. For example, for the Centre Pompidou in Metz (by Shigeru Ban), what we got basically was a triangular mesh. What we need as a base is a continuous surface. So if it´s very complex, like in that case, we have some other people who usually work in the automotive industry to help us creating this design surface. And they have their own kind of software tools to create this high quality surfaces. If you do a reflection test it is really smooth everywhere, there are no kinks. If you have a patched surface that is made of different parts, you will always have a problem with the seams. If you project something on to it, you get choppy lines or curves, so before you go on you have to join them. So it is a very good idea to have in the beginning a high quality tridimensional data.

You have to take under consideration that (under our office in Stuttgart) our clients are architects. They come rather early, maybe with just a competition model, which is anyway very rough. In those cases we develop this basic geometric model together, so that is a smooth thing, which goes to a certain extent into the production.

But (at the Stuttgart office) we wouldn´t do a 5 or 6-milling Gcode, which is something that Fabian (who runs the office in Zurich) does. There, we just get the finished models from our clients, and this is usually problematic. First because of the quality issue, but also the detailing sometimes – it may work in the CAD model but not in the reality. We usually have assemble problems. Of course you can draw it and everything fits together, but there´s no way to get it together. So this is also part of our work in collaboration with structural engineers. This is what we call sometimes "normalize" this design.

MS: When you need to make a change in the design, how do you proceed?

AW: There are different types of architects. Some of them are very strict, for example the Italian architect Maximiliano Fuksas, they do some kind of form and they want exactly this to be built. Of course I don´t want to change the design, but I want to find a good balance between aesthetical aspects, aspects of structural design and aspects of production to make it kind of reasonable in the overall performance. Of course you can theoretically do any kind of form, but I have to admit that I wouldn´t do certain jobs. For example Frank Ghery, from my point of view his work is more sculptural. There is not this general line or what we call in German a red thread (an underlying logic), it is just doing things randomly and this has an effect on how the projects are worked on. Usually these projects are done in CATIA (2), where you can still use certain aspects of parametric modeling and associativity, but it´s not a fluent design idea, it´s more a lot of features packed together. Of course it´s a very expensive way of doing architecture, although it has certain qualities. I think you could never really script in an effective way, but there are other tools in order to come along with these designs.

MS: What kind of software do you use?

AW: We are interested in software which just have a lot of geometrical features, functions and stuff like that, and they should be programmable. So if you need the intersection line between two double curved surfaces in space, I would never be able to set up the mathematical basics to get that, but there´s a function to do that, and this is what we need. With these basic tools we put together the design.

There are good architectural packages, with all the buttons and all the libraries and you can do the architecture you want. But you have to know that you are limited. I don´t want to be limited by the tool at all. Therefore, in between there is this exception, for example if you look at a stair. If you want to program a stair macro which can do any staircase, somebody would think about with different setups. It would be so complex, first to program and later to use, because for a lot of people there would be a lot of options that they would not know what they are good for. Therefore, this is not the way to do that. When you go away from this standard, then you see it´s better that you write your own script for your specific stair. This is not complicated, you get exactly what you want and you are totally in control. I don´t want to be limited, because for me standard software tools are always conservative and old fashioned. They take a long time to be developed and it takes a long time before new influences are part of it. I already understood that even in the eighties. At that time in Berkeley they were already working on tools to optimize floor layouts, space layouts, but I don´t know any CAD package that have that tool incorporated - why?

GC: Do you think it would be too boring to have the computer to do that for you? There is something else rather than the optimization?

AW: Maybe it´s not possible... a lot of people say the computer would do the design. I don´t believe that ever is going to come out something interesting. Maybe you find some kind of an algorithm that does something. But it´s always you, you direct the process. The computer for me is a tool, and you are responsible for the results because you develop the algorithm. It´s your decision. Even if you don´t understand the algorithm you still choose it.

MS: And what are the challenges of working with industries that are not used to producing building parts?

AW: Normally, we are coming up with designs that result in an unusual structure where the normal building industry is not prepared for. You don´t have a company which have done that before, so you have to look where can you find a company which is close to this or at least has the right machines or has experience with the material. Sometimes we have companies that are perfect to produce the design but they have no staff for the assembly on site, normally they deliver the product and it´s another people’s problem to assemble. A good thing is to work with façade companies, because they are usually quite experienced in dealing with glass, steel and metal, and they have experience in assembling things on site. But is always a bit of a looking around, and a very important aspect for us is that we feel that the company is interested and motivated. Some companies say ok let´s do it, but they don´t stop to think about the project before they don´t have the contract and this is a big risk for the client. Because once they get the contract and start thinking about the project, they might find out it´s more complicated than they thought, then you might have financial problems and so on. There´s a big responsibility in choosing the right partners. You have to really trust the company and believe that they are really motivated to bring to an end to all of this.

GC: What is your role in choosing the company for the client? Would you suggest a company or say this can only be done with that? Do they have to go bidding?

AW: You are right, I do make suggestions. I might say to the client: you better take this company, because if you don´t take it I´m not going to work in this project anymore. Because I know what can happen. I´m just assuring quality for my client and wouldn´t allow this kind of companies "nice project, let´s try it, it´s good for the pictures". That´s not enough. A lot of those industries, although they have the capabilities, they are not used to dealing with freeform shapes; they just don´t have experience enough. We very often go quite far in terms of proposing how to produce in an effective way. On the other hand we are aware that we are not able to take the final responsibility. We propose that to the company and we have the feeling that they take that and even develop it further on, and then we are very satisfied and very calm. We find a partner that we can work with, and then they add their expertise, because we are still only architects, another specialist in certain way. I can go to certain extent but then I have to have the feeling that I can give the problem in responsible hands and they will continue working on it.

MS: What makes a complex design be more expensive to produce?

AW: If you can do it by digital production it should not be really more expensive. It´s difficult to say because when the design is more complex, it means you can´t do it with standard right angles. You can find a way to produce it which is way out too expensive because you didn´t think about certain things or you drive people into problems which are not expected from them and this could really harm a project. If you are an insider, you know that a lot of this projects haven´t worked in the way they were planned. A lot of them took years to finish, the cost was too high. I can say from the projects we are in, we can really avoid this. The Mercedes Museum and the façade in Davos were delivered in time and in the expected budget. In the case of the museum, at least one company which was involved didn´t really make any profit, but that sometimes happens as well. But we are very satisfied when the company does it and earn their money and the product is fine. Sometimes you have to say: no, I don´t want to be involved in this, I don´t believe it´s going to be successful. 

MS: Do you discuss about materials and their effects with your clients?

AW: There are two aspects, first it is just the visual. Somebody needs one certain metal looking surface. It´s very rare that there are structural aspects, in terms that you are forced to take certain material, because otherwise it wouldn´t work. The façade in Davos was, for a long time, though as being done in aluminum, because we thought it was lighter and also it could have a structural effect. But in the end they chose steel because the material is much cheaper than aluminum, and this had another effect. When the façade was already planed and in production they increased the thickness of the plates by a few milimeters, which caused a lot of additional problems for the concrete people, who had to add more concrete and more reinforcement. I don´t know if this decision was really made in order to save money. They just shifted money from the façade into the concrete part, which could be for political reasons, could be also a strategy, I don´t know how it worked. Another thing is that usually at the beginning of the project money doesn´t count. I have seen a lot of projects which started with carbon fiber, and it ended up with black steel (standard steel profiles). A lot of people would like to have these fancy materials, of course in terms of cost and production there are not so many companies, especially in the building sector, which we are used to use those materials. If you go, let´s say, in the sailing business it´s just a totally different thing.

notes

1
Rhinoceros is a CAD software for 3D modeling.

2
CATIA is a CAD software for 3D modeling.