Digital Design - Experiment in Contemporary Architecture

DIGITAL DESIGN - EXPERIMENT IN CONTEMPORARY ARCHITECTURE Andra Panait Ion Mincu University of Architecture and Urbanism, Bucharest, Romania andra_panait@yahoo.com Abstract The unprecedented contemporary expansion of knowledge brought forward a requirement to adopt a new mentality and a sense of urgency in bridging disciplines which, by the middle of twentieth century, gave rise to multidisciplinarity and interdisciplinarity. Mankind has always used intelligence to address problems that life and society have generated. The same planet with the same resources needed to support the life of an exponentially growing population. In response to these requirements, the science and the technology have also grown exponentially, giving rise to cybernetics, mathematical theories that aim to translate the nature at macro and microscopic levels and theories that had the potential of being useful in generating new solutions in architecture. The emergence of computers, of digital technology, of complex graphics and visualization techniques streamlined the aproach to architectural solutions and proved useful in satisfying both the pragmatic constraints that make architecture a science and the features that make it an art. This paper investigates the design of digital processes and how they can improve the thinking process at the project concept stage and the rethinking of the relationship between architecture, sciences and, inevitably, a highly technological society. Also present is a subtextual issue as to what extent these current trends are found in Romanian schools of architecture. Keywords: technology, interdisciplinarity 1 digital process, parametric, algorithmic design, experimentation, OVERVIEW In the past centuries, the evolution of technology has become the main engine of urban transformation, in all its aspects. The speed made the space relative and expanded the city’s flows by relating it more and more intensely to the environment. This generated new complex processes of restructuring the urban existence by a new logic of the connection and use of the information for an informational society. The physical and virtual systems that underlie our society cannot be managed except through latest technology. [1] The computer and the internet have transformed the perception of space and are synonymous with the development of a new type of informational society [2]. Although it shows more visible results in the recent years, the digital revolution has its roots in a relatively long history. The massive expansion th of the information from the early 20 century or the intensive use of the computer for the simulations during the Cold war represent the key points that lately had technological consequences for the digital revolution within the architectural area, design and urban planning. There is a major advance in the practice of the architecture due to the use of informational technologies. New tools have enabled architects to produce more exact drawings, faster and easier. Computers offer an undoubted support during the design phase and for the composition of the construction documentation. However, what happens to the first two phases, of program and schematic design – essential phases of the concept? Most of the design software, rather than assist the architect during the design stage, are built on the assumption that he passed the experimental stage of ideas and has already reached a solution. Can the digital environments facilitate a better conversation between the architect and his ideas? Moreover, during the concept phase the architect uses a series of tools, methods and different environments and during the transition from one environment to another a part of the information can be lost. Can the new technologies link the analog and digital, the 2D and the three-dimensional; can they enhance the design ability of the architect? I think the answer is positive. There is a great potential for the new computational technologies to provide advanced tools that would better serve the needs of the architect during the concept phase and, ultimately, to contribute in a positive manner to the quality of design. The study focuses on cases where adopting various methods is not only intensive, but removes technology as a concern and transforms the technological aspect into a tool. More specifically, the only contribution of technology is that it facilitates the study of principles and phenomena, which exist independent of technology. Technology only provides the computing power without which these studies would be very laborious or impossible. These concerns gradually enter the school programs (in Romania timidly and late) with the aim of providing the future generations of architects the necessary mental framework to create a synergy between the complexity of the computing methods and the creative use of the computer and to give them the ability to adapt to new ideas, necessary in an area that appears to be changing so fast. 2 METODOLOGY This paper aims towards an interdisciplinary approach making a parallel between the architecture and other areas of inspiration where the architects inspire their ideas from: biology, mathematics, computing etc.; some of these methods are then illustrated on several studio projects exploring the interaction between methodology, technique and representation and highlighting the influence that digital environments have upon the conceptual process. They also explore various geometries related to specific forms such as topological surfaces or independent structures acting together in a larger system. This paper makes a brief passage from tools which, while helping the architect to express his ideas are not directly involved in generating the concept, to tools attempting to assist the initial process of exploration. The cycle of exploration, evaluation and refinement on which the human process of creating a design is based is, essentially, evolutionary, so naturally it led to the analogy with the algorithms underlying the evolution and natural selection. This algorithmic step is the one that uses the borrow from other areas in turn influenced by the availability of the computing power. These borrowings are made by taking the algorithms and methods of study in those fields. Searching the appropriate algorithms, is still in its beginning but the prevailing algorithms are from several categories [3]: computational geometry, rule-based systems, selfgoverning systems and optimization, which will be detailed and extensively illustrated through projects. 3 THE EXTENDED TERRIORIES OF ARCHITECTURE Over the time, both the definition and architecture’s fields of interest have expanded and passed through important moments. Although not as prominent as the explicit inclusion in the XV century by Leon Battista Alberti of the entire built environment within a definition of architecture, we see today how architecture, even seen only as a way of cenceiving the built environment, is an interdisciplinary field, connected to the latest scientific methods. We are witnessing not only a diversification of the fields of inspiration - biology, mathematics, computing – but also to a change of the way in which this influence occurs in the act of conceptualization. In 2002, through an article entitled “The New Paradigm in Architecture” [4] Charles Jencks announced a new way to make architecture, basing it on a series of scientific theories as the theory of complexity, self-governing systems, fractals, non-linear dynamics, and emergence. Although other commentators, with Nikos Salingaros [5] being the most vehement, emphasized some confusions and syncope of the rigor attacking both the idea as well as its examples, the years since then have really seen an architecture strikingly different. It was not necessaryly produced by the names mentioned in the article but certainly proved that the architecture can be extended in areas that at first glance do not provide an obvious connection. There are combinations of architecture and landscape, biology, mathematics, computing which produce new types of models outside the scope of pure architecture, and form an internal study of the language of the architecture in collaboration with external approaches to these fields of influence. This expanded field of the architecture is mostly due to the experience gained in the fields of the other sciences, in turn driven by technological development. Since about 10-15 years, the bio prefix prevails in the general scientifical research and everything that is related to it: engineering, genetic engineering, technique, structure, mimetics, etc. The present stage follows that of 1975-1990, in which automation was dominating and everything connected with it, as well as during 1955-1975 a proeminent focus of researches was the fluid mechanics. So actually there is no randomness in the challenge that comes from the disciplines of life, from the organic world, higher or lower organized, and which is assumed by theorist architects and applications. Neutral networks are mathematical models for the process of learning in the brain. The fuzzy logic imitates the human brain in the process of making decisions. The genetic algorithms imitate the mechanism of evolution and natural selection and it is an almost universal way to search for optimal solutions. Cellular automata simulate the growth processes. If architects turn to the computer and the mathematical models in the design process, it is inevitable for them to connect to what happens in the areas adjacent to their discipline. There are also detractors of the new, and it’s not surprising that one or the other repudiates the idea of using the computer in the architectural design. An attitude hard to understand, especially when it’s about a technical procedure, searching for the optimum, as in the project described in Chapter 4, in which the algorithms were used to minimize a cost index of an apartment building in Manhattan. I think, however, that of interest is not the purely technical framework but the one which provides meaning, purpose, symbol and creative value. Investigating the computation concepts and their importance in digital design raises the value of netrually the computer as a tool to a dynamic way to approach the design process, which leaves a mark on the final result. As in the case of mathematical functions, the defining elements of a calculation function represent an algorithmic relationship, well defined and measured between two sets, in our case the set of input parameters and the set of output variables that have guided the definition in the digital design of three types of levels [6]:that of representation, parametric and algorithmic, summarized below. 4 FROM REPRESENTATION TO ALGORITHMS There is a gradation of technological influence on the design process: The first step was – the most obvious – digitizing the design. In 1963 Ivan Sutherland [7] presented the first CAD program, setting a model which remained unchanged until today. In 2008, a greater calculation power allows a more sophisticated drawing. But the level of use is still the same – the representation one. So, this grading is not about chronology, but the way in which technology is used. Fig. 1 - Ivan Sutherland, Sketchpad, First CAD Program 1963 and 2008 (source: http://www.cadazz.com/cad-software-Sketchpad.htm) The next step is that in which to a predetermined geometry there are applied various operations: deformation, twist, spin. A deeper level of use is the parametric one, in which there is no predefined form but one expressed as a set of parameters and expressions, that is, relationships between parameters. The parametric approach builds into a more complex stage, which can be viewed as a generalization of the parametric one: the algorithmic stage. Forms are generated by algorithms, working not with the actual forms but with “recipes” describing their generation. This algorithmic stage is the one that borrows from other fields, in turn, influenced by the availability of increased computing power. These borrowings are made by taking the algorithms and the methods of study from those fields. The algorithmic architecture is trying to express through code the rules and principles of an architectural typology. This expression is reflected in the design of programs capable of generating spaces and forms. The search for suitable algorithms is still in its infancy, but the algorithms from some categories prevail: computational geometry, rule-based systems, self-governing systems, and optimization. A classification of the main algorithms used in architecture (Dr. Toni Kotnik, ETH Zurich) • Computational Geometry: Voronoi diagrams, pathfinding Algorithm A* (A star) • Shape grammar Rule-based systems: L systems • Self-governing systems: Cellular automata, Swarms • Optimization: Genetic Algorithms I will examine in turn, some examples from these categories: 4.1.1 Voronoi Diagram Fig. 2 - Voronoi Diagram in Nature (Dragonfly Wing Pattern and Sun Flower) Described by the mathematician Georgy Voronoyi, the diagrams that bear his name represent a partition of the space into regions. Starting from an initial set of points, a number of cells is generated. Each point of such cell is closer to the point that generated the cell than to the other points of the initial set. A structure similar similar to a living tissue is otained, with the points of the initial set representing the nuclei of the cells. [8] There are many examples in nature that follow the same geometric rule and which, in mathematics, is called “voronoi growth model”; for example the sunflower seed distribution can be simulated with a perforated grid. Beyond the possibility of using them as a decorative mechanism, these processes are useful for their structural properties into the two or three-dimensional space or as a way of organizing the space based on the neighborhood or proximity relations. The next example is a school project of collective housing – 80 apartments + commercial spaces + community spaces + leisure facilities – in which voronoi partitions were searched that would meet the program’s requirements: for example, surface, depth, distance between buildings, site organization. A system of independent cells which still work together as a larger system was obtained.. Fig. 3 Dimitrie Ştefănescu and Veronica Popescu Collective Housing Project, 3rd year, Architectural Design Studio (source: http://dimitrie.wordpress.com/architectural-projects/) The project authors are two students: Dimitrie Stefanescu and Veronica Popescu. The first chart represents the view of the grasshopper definition that led to the generation of the solution. They started from a voronoi algorithm to create a series of exterior and interior courtyards with different degrees of intimacy that make the transition from public to private. Using the parametric methods made it possible to control the theme data – the built area, floor height, sun exposure time, the ratio between the commercial and housing areas and the leisure one, etc. The work included a series of shadow studies which also led to the overall compozition. Although the borrowing in this case is rather formal, the overall compozition of the project achieved a certain quality of the spatial solution, especially on the urban level, a good grading of the public and half-public spaces and an interesting volumetry. 4.1.2 L Systems An L-system [9] is a particular type of dynamical and symbolical system generated by a formal grammar. It was originally designed to model the growth processes of plants. An important quality of the L-systems is that, starting from a small number of rules they can describe complex shapes. The first stage of an L-system is choosing the set of characters to operate with and establish a replacement rule. It is an abstract level, in which all work is done only on characters. Starting from an initial letter and a replacement rule, successive replacements are applied. These character strings may be assigned meanings. For example, “A” can mean “draw a line”, “+” could mean “set the angle for the next line” etc. Fig. 4 ‘Weeds', generated using an L-system in 3D and Sierpinski Triangle. (source: http://en.wikipedia.org/wiki/L-system) Of interest here is the attachment of more complex meanings, that signify certain geometries, or which guide a growth, preventing, for example, through simple rules reaching below the ground. The Lsystems are thus a mechanism through which, starting from abstract expressed rules, complex and coherent forms can be generated. Another example, also from the projects undertaken in the workshop – isolated home – belonging to the student Dimitrie Stefanescu. Staring from Menger’s Cube, he did studies of algorithmic control of the surface’s porosity. The purpose was to create an adaptive environment, porous, which would allow exchange of air, sun, perspective controlled points, without losing the family privacy. Fig. 5 Dimitrie Ştefănescu Individual Housing Project, 2rd year, Architectural Design Studio (source: http://dimitrie.wordpress.com/architectural-projects/) 4.1.3 Genetic Algorithms Evolutionary architecture sees the built environment as a form of artificial life, applying to it the principles of genetic coding, replication and selection. Among others, it starts from observing the discrepancy between the existing balance in the natural environment and the disruption it causes the built environment and suggests the possibility to bring to the latter something from this balance. [10] At the heart of the natural selection there are three parts: 1). The Genotype, that is, the “building instructions” – genes / heredity. 2). The Environmental action. 3). The Phenotype: the sum of characteristics of an individual that is the result of the interaction between genotype and environmenT. The adapted phenotypes survive and their heredity is combined in order to obtain new phenotypes. Fig. 6 Natural versus Built environment For the application of genetic algorithms in architecture, the architectural concept must be expressed as a (genetic) code. This code is mutated and developed by a computer program in a series of models in response to a simulated environment. These models are evaluated in the environment and the code of the "successful" models is used to generate a new cycle, until a certain stage of development contains acceptable solutions. To obtain an evolutionary model it is necessary to define: a genetic code, the nature of the virtual environment and the selection criteria. The obvious difficulty is to define such a code in terms of architecture and therefore this is a fertile ground for experimentation. In the following example an algorithm generates and evaluates a building, or – more specifically – its circulation node configuration and its apartments based on a fix set of criteria: light, view, noise factors, maximal surface. This last criterion is useful, moreover, to emphasize how careful the establishing of this code must be. For example, in this case, the author overlooked the fact that optimizing a building for a maximum enclosure surface basically optimizes it for maximum energy consumption. The project belongs to Michael Hansmeyer and aims to develop an apartment building containing units of different size, using optimization by genetic algorithms. It starts from the question: What is the role of the architect in the apartment market nowadays? With few exceptions, apartment buildings are identical concrete structures, with minor variations applied to the closing walls of the facades. In most cases the only difference is the internal partitioning; inner spaces – separated both by structure and façade – are set in the last minute to best address the market requirements of the moment. The project addresses this situation by transforming the architecture in a pure exercise of space partitioning. The building’s design is reduced to a problem of optimization, in which the maximum area of the site’s enclosure is segmented for an optimal configuration of the apartments. The optimum is seen as the maximum value on the market that this segmentation can achieve. Design process’ flow 1. The input: • Rules of construction: Placing the traffic node, corridors and apartments • Site characteristics: Maximal surface, light, view, noise factor • Apartment characteristics: The catalog of apartment types with dimensions • Rules for assessing the apartment price: The apartment price increase sensitivity to factors such as the landscape or the height that it is located on. 2. Computation • Construction process: The node placement; corridors positioning; units positioning; • The assessment process; calculation of the market value of the building based on the evaluation of each apartment. 3. The output • Construction scripts: visualization of the variants in the CAD program (in this case, Maya) • Construction specifications: the detailed list of all the apartments and attributes • The market value of various options The aim in using this algorithm was to design the building so that its market value is maximized. The algorithm recursively generates and evaluates the building, requiring approx. 40,000 iterations until it reaches an “almost optimal” point. The algorithm produces not only market value, which is the main objective and which acts as a reference point for the next iterations, but also provides detailed specifications that are inputs for the next secondary algorithms – such as calculating a structural system. While the optimization leads to a generation of a built form, the outline is limited by rules of construction and their parametric values, which in this case, prescribe a set of predefined units. Fig. 7 Optimization Study for an Apartment Block in Manhattan, Michael Hansmeyer Project (source: http://www.michael-hansmeyer.com) 5 THE RISK OF FORMAL BORROWINGS. METAPHORS VERSUS SCIENTIFIC DISCOURSE Searching some algorithms able to generate shapes well adapted to the natural environment inevitably turned the attention towards adaptation mechanisms found in nature, constant source of inspiration throughout the history. This “return to nature” was also one of Charles Jencks’ visions, who saw the end of the century as marking an evolution of culture towards a more complex interpretation based on biology and mathematics [11]. But a change occurs on the level where this takes place: not from the forms of nature but its processes. We see how this extension of the architecture into new territories creates both a scientific discourse as well as a metaphorical one, but the line between them must be treated with caution. The strange mutations of the scientific discourse, when it is rewritten in architecture, mark it intrinsically. This becomes an area where the external ideas are not only imported but formulate new theories, where architecture’s technologies realign their material according to their own specific laws. Even if the terms of the theory of evolution, or the slang of any other field, had been emptied of their complex content when they reach the architectural discourse, this does not make them uninteresting or useless; at most they cast a shadow on the intellectual honesty of those who use them to try to give scientific legitimacy to a specific approach and do not take care of specifying the metaphorical intention. It should be noted that this “water muddying” is not necessary - analogies can also be the starting point of a creative process. They imagine new relationships between apparently dissonant materials, visualize an idea or an operation, they make it known as to be investigated later, eventually becoming the starting point of a creative work. On the other hand, in the approach of a scientific discourse and the assuming of new rigorous methods it is desirable to not abuse the presumption of honesty. Some descriptions of the processes that led to new forms and solutions suggest the necessity of the introduction into the architectural education of explicit information that the scientific activity is guided by specific methods and that a scientific process means to participate into a community and provide all necessary data for the independent confirmation of the results of an experiment. The experiment (from Latin ex – and periri “about trying”) is a set of systematic observations developed in the context of solving some particular problems or issue, to support or refute an hypothesis or research concerning phenomena, being a fundamental tool in the empirical reaearch. Physical experiments always served the architects as a starting point in investigating the design. Well known examples are the structural model of Gaudi de la Colònia Güell from Barcelona, that consist of chains from which there were hung weights, Heinz Isler’s membranes or Frei Otto’s tent structures, the latter constantly stressing in his speeches the importance of the experiment in the practice of architecture [12]. His so called natural constructions made it possible to extract structural principles that could be then implemented in the architectural ideas or in solving constructive problems. The purpose of these experiments was to mathematically represent the surface. Combining the empirical approach with the theoretical and mathematical one useful conclusions could be drawn from them and, therefore, are welcomed in the pedagogical activity. Fig. 8 Gaudi Model. This very closely represents the finite element method in a physical way. (left) and Heinz Isler’s Membrane (right) The experiment, as a research method, fits itself in all methods of a science, but it cannot remain isolated from other research methods. Given that in the field of architecture we deal with a number of theoretical concepts, the experimental method can be interwoven with them and the result filtered through the prism of all other factors that may influence a solution. Otherwise, the result cannot gain legitimacy simply because at its base there is a scientific method. For the architect, science is necessary both in terms of its practical spirit as well as in terms of a way of knowing along with other revealing forms. Probably that living in a city emptied of mathematical models that are found in nature and old buildings is not the main cause of the lack of interest in mathematics and for a certain scientific rigor in some schools of architecture, but rather a change of mind, the desire to quickly have a spectacular result that would shock is more important than the fulfillment through knowledge, which requires time and effort. 6 CONCLUSION Starting with the technological progress, more aspects of human activity fall under the explosion of the computing technology. Claiming itself as 'the most social of the arts” architecture is a visible expression of how the society gets transformed. It has a social dimension shown by how it materializes – first in programs then through materialization and physical expression – important aspects of life and human activities. Most of the times these events, whether theoretical or physical, use interpretations derived from the technology and contemporary values of a given moment. Informatics has become increasingly interwoven with many sides of life, to be a part of the daily life. The high degree of accessibility made the computers replace some crafts but especially to substantially modify the way others are practiced. Architecture not only makes no exception but it is subject to influences manifested both directly, through using informatics in the design practice and indirectly, as a mirror of the society that is still searching for the mechanisms to control the ways in which the public and private life are affected by an omnipresent technology. We saw how the transformation of the society and the developments in other fields due to the availability of increased computing power can put a mark on the process of thinking an architectural project. The effective use of these concepts involves absorbing a volume of knowledge and requires a more rigorous and abstract way of thinking. Increasingly, architects learn and explore new ways to mathematically define surfaces and geometries represented by functions which describe a range of possibilities. They exceed the condition of clients of a company that aims only to market a product of CAD or modeling and they train on their own, or with the help of the programmer and computer, and they try to create specialized programs to meet their vision and help them express it. They are, besides scientists, engineers and artists, component of the increasingly wider inter-digital field, where the information generates a dynamic process of change in many domains. The projects carried on by students in the workshop constitute an example of influencing the architectural thinking, in the stage of conception, by the grammar and logic of the algorithmic language inspired from the latest scientific research. We see how these new concepts emphasize even more the prevalence of the process over the product and an implicit intention to validate the final result in terms of the processes used. However, the process can at most provide accuracy to a solution, but it does not confer necessarily architectural value; maybe it should therefore remain in the background, and the criteria that distinguish a valuable building from a regular one should be related more to the qualities of the object itself and its integration in the context, than to the technique used in its design process. The reality is that currently, in school, the computer is commonly used just for the primary level of representation and there are rare the cases where it is involved in the process of design and this approach is rather done on an experimental level. On the other hand, there are only a few cases in which these experiments are being theoretically supported the rest being only formal, without a conceptual justification or economically inefficient or unsustainable. Even if the experiments are supported by a theory that is valid in other fields, the criteria by which such a theory is chosen must be rationally justified through the aesthetic, functional, technical, economic qualities that it can provide. I would like to emphasize that all the methods described, in their substance, have nothing to do with the computer. Incidentally, in the present time, the environment that facilitates the exploration of these concepts is the digital one, because it allows the integration of a large number of information and methods in the process of design. Obviously, the final arbiter of the validity of a solution is the building and reassessment as a result of the interaction with the community which it is placed into. REFERENCES [1] Metapolis Dictionary of Advanced Architecture: City, Technology and Society in the Information Age, Authors: Eduard Bru, Jose Alfonso Ballesteros, Stan Allen, Cecil Balmond, Marie Ange Brayer, Manuel Delgado, Jose Miguel Iribas, Jose Morales , Willy Muller, Markus Novak, Fernando Porras, Federico Soriano, Mark Wigley, Ole Bouman, Aaron Betsky, Inaki balos, Karl Chu, Juan Herreros , Xavier Costa, Manuel Gausa, Vicente Guallart, Willy Müller, Actar, August 2003 [2] Weibel Peter, Virtual worlds: The Emperor’s new bodies, The MIT Press, Cambridge, 1999. [3] Toni Kotnik, Algorithmic Architecture, Introduction to the MAS Colloquia 2006/2007 [4] Charles Jencks, The New Paradigm in Architecture: The Language of Postmodernism, Yale University Press, August 2002 [5] Nikos Salingaros, Anti-Architecture and Deconstruction, 2004, ed. ISI Distributed Titles [6] Toni Kotnik: Digital Architectural Design as Exploration of Computable Functions, International Journal of Architectural Computing, Vol.8, Nr.1, 1-16, 2010 [7] Sutherland, Ivan Edward (January 1963). "Sketchpad: A man-machine graphical communication system (courtesy Computer Laboratory, University of Cambridge UCAM-CL-TR-574 September 2003)". Massachusetts Institute of Technology. Retrieved 2006-12-26 [8] Franz Aurenhammer (1991). Voronoi Diagrams – A Survey of a Fundamental Geometric Data Structure. ACM Computing Surveys, 23(3):345–405, 1991 [9] Grzegorz Rozenberg, Arto Salomaa - Lindenmayer Systems: Impacts on Theoretical Computer Science, Computer Graphics, and Developmental Biology ISBN 978-3-540-55320-5 [10] John Frazer, An Evolutionary Architecture, Architectural Association Publications, 1995 [11] Charles Jencks, The New Paradigm in Architecture: The Language of Postmodernism, Yale University Press, August 2002 [12] Toni Kotnik: Experiment as Design Method Integrating the Methodology of the Natural Sciences in Architecture, International Journal of Architectural Computing