Synthetic Computational Design

Synthetic Computational Design



Gao Yan


“The designer finds himself in a specific historical situation, facing some particular problem. He responds to the logic of that situation with some design solution, and this in itself produces a change in the problem: it creates a new problem.”


Philip Steadman



In the twenty-first century, new terms increasingly arise, with which to describe new ways of thinking and operating based on developments of new technologies, which can be understood as the sum of technical theories. ii The application of digital design and manufacturing tools has profoundly changed conventional design methods and building processes. Parametric and computational approaches have emerged as alongside sustainable design, and interestingly, these two trends have only recently begun to become correlated. The first generation of digital architects, such as Greg Lynn, Frank Gehry, Zaha Hadid, and several others, opened up a new territory by embracing new digital design methods. A second generation continued to explore the possibilities of these new digital tools for architectural design, while the current generation of countless young designers is working towards the exhaustion of computation as new design problems are being created.


The first and second generations have demonstrated some of the possibilities of these new methods, in how we can technically build anything, as long as we can imagine it. If this is indeed a dream come true, then a question to ask the current generation is what shall we design if anything can be built? If past generations had committed themselves to the pursuit of the New, the current generation ought to aim for greater synthesis in their propositions.


What is synthetic computational design?


In order to articulate a definition of synthetic computational design, we should firstly delineate approaches to computational design which are not classed as synthetic. Non-synthetic design focuses on highly limited scopes of the entire design process, for example, aiming solely to invent new forms, to create new types, to explore new prototypes. It focuses on fragmented stages of the whole, as a single layer rather than multiple layers of a synthetic approach to computational design. Non-synthetic design severs the complex connective network of the mechanisms informing an architectural project into isolated tasks. Synthetic computational design covers a broader range of consideration throughout computationally driven project, including not only formal and spatial ideas, but also communication and procurement; not only generation and automation, but also optimization and evaluation; not only quantity and quality control, but also risks and costs management; not only aesthetics, but also efficiency; not only durability, but also flexibility; not only innovation, but also fitness. It is essential to map out these complex workflow networks, and to engage holistic processes with cutting edge computational tools rather than concentrating singularly on particular, isolated aspects of project workflow.


Consideration of a broader scope of processes is not enough. Distributed knowledge across different disciplines must be integrated coherently through computational means. The translation of project information into a quantified database shared by the design team, aims to distinguish, as well as integrate, the criteria and goals of a project. Design decisions are made, negotiated, and evaluated by all relevant parties, instead of being determined in a fragmented mode and then collided.


Thirdly, synthetic computational design must be meaningful for society and the environment. The measure of significance has many scales. Environmental significance can be measured through well established sustainable design evaluation systems, such LEED or BREEME accreditation. Here I am more concerned with social significance than with over-familiar environmental arguments. Synthetic design processes should not simply be driven by the strong desires of a singular designer, but rather being in correspondence to an intricate understanding of context, and in return, creating new contextual conditions with which to inform the design process. Architectural form can always be understood as the expression of culture and context, and at the extreme, architecture inevitably becomes symbolic. “Restoring symbolic meaning is a most fundamental task in a metropolitan world in crisis of communication. … Architecture, of all kinds, must be called to the rescue in order to recreate symbolic meaning in the metropolitan region, marking places in the space of flows. …” iii Symbolic, or rather, iconic architecture, I would argue, should not be criticized because of the deployment of symbolic form, but rather for the lack of integration to existing public space. Architectural form can increasingly be understood as cultural expression in the Information Age. Computation offers designers powerful tools with which to innovate and, to go beyond subjective inertia of individual styles. It also propels design away from the superficial mimicry of form into a new territory of understanding the logic of formation, so that architectural knowledge can be embedded into metaphoric forms. Emergent forms can be generated and negotiated with designers’ intentions under specific constrains. Designers should learn from the process of engaging with feedback mechanisms which reshape initial intentions. In other words, synthetic computational design enables more meaningful iconic architecture.


Why synthetic design?


Reductive ideologies and methodologies dominated architecture in the twentieth century, perpetuating the belief that by understanding the parts of a system, the entire system can be comprehended. Avant-garde practices often engage in focused, isolated aspects of architectural design, to facilitate experimentation with new techniques. Experimental design is therefore not normally synthetic, as it focuses on particular aspects of a complex project in order to manage innovative thinking, with the hope of delivering the New. Nevertheless, experimental design is necessary to proceed to synthetic computational design, because, without understanding the parts, there is no way we can understand the whole.


Today’s world is inter-connected, and multiplicities of relationships and interactions are more significant than singularities. By no means can we fully comprehend all the relevant issues of an architectural project by breaking them down into pieces, nor can we propose a meaningful scheme which exhausts the every problem within a context. We must, however, prepare ourselves to engage the whole with collective cooperation of distributed intelligence through computational platforms in order to maximize the effectiveness and efficiency of a design scheme. A parallel analogy can be made with Western and Eastern approaches to medicine, where the former directly targets the locus of pain, whereas the latter focuses on the micro-network causing the pain. The consequences of the Western approach are immediate effects but with more side effects and less durability, whereas the Eastern approach to medecine cures slowly but may have less side effects and longer relief. The Eastern approach aims for long term effects and is therefore more durable, as does synthetic design, which understands problems of an architectural project as a network of interactions.


Many people still regard computational design as solely creating expressive and expensive forms, rendering judgments from a superficial level, disregarding irregular forms, but failing to understand the significance to design processes and discourses, modes of manufacturing, and the qualities of computationally driven products. Computational design has evolved from obsessions with the creation of forms, towards intelligence, efficiency and automation throughout the entire life of architectural constructs. Synthetic approaches are gradually stitched together by increasing interests in various aspects of architectural industry rather than creativity being limited to formal and spatial issues.


Synthetic computational design targeting sustainable development


No doubt sustainability has become a central priority in the twenty-first century, aiming for patterns of resource use which aim to meet human needs while preserving the environment, so that these needs can be met not only in the present, but also for future generations. Sustainability will inevitably become ubiquitous, as will computational design. The underlying nature of computational design is optimism, corresponding to the core values of sustainable development, which aim to minimise resources and maximise performance. In this sense, the time has come to target sustainable design with computational design systems. Sustainability stands as intentions, while computation substantiates the technical means to confront these challenges.


Sustainable development comprises three dimensions: environmental sustainability, economic sustainability and socio-political sustainability. In short, environmental sustainability concerns the impact to the ecological environment by balancing the consumption of resources and deterioration to the eco-system caused by human activities; economic sustainability encompasses opportunities for economic growth which treats nature as economic externality; socio-political sustainability refers to the equal or greater access to social resources as the current generation. These three dimensions should be integrated into agendas of sustainable development to induce more holistic approaches to the building industry, which contributes to nearly 40% of global CO2 emissions, and vast quantities of raw material consumption.


Sustainable architectural design has been addressed for many years, to the extent that its principles and constraints have been codified into building regulations in several Europe countries. Sustainable building products have been developed and are proliferating in the past decade. There are, however, three fundamental problems associated with most sustainable design practices.


Problem 1: Fragmented design approaches


Only when the social, economic and ecologic aspects of a project are integrated, can a coherent design solution be achieved. Otherwise, projects claiming to be sustainable remain as unrealistic show-pieces of so-called sustainable architecture. These claims may work in the climate this decade, but they may also become superseded by conflicting and changing interests of local communities and economic fluctuations. A holistic design approach capable of tackling sustainable development must be adopted to balance these three dimensions of sustainability. The greatest opportunity to achieve genuine sustainable design lies in the integration of all design factors into quantified data generated and managed through computational platforms.


Problem 2: Lack of inherent thinking of spatial and geometrical aspects of architecture


Due to the materiality of architecture, most efforts have been invested on building materials, energy preservation and supply, and construction methods for the sake of sustainable development. Contemporary design practice can be reduced to the selection of the standard products and placing them into a framework. An extreme symptom of this mode of practice is when architects design forms and spaces with neutral materials which are then filled by material suppliers. This leads to retrospective design modifications to cope with the material constrains, and therefore this workflow mode wastes time and material resources.


Architects are no longer masters of building materials, due to the overwhelming amount of resources, where new products flood onto the market every day, and nor are they the masters of building technologies, which are increasingly complex and distributed. What remains for architects is to be the master of form and space driven by all relevant parameters. Can architects address sustainable issues when they draw their first sketches as a mode of geometric expression? Instead of engaging in processes from design generation towards optimization, can optimization happen early in the design process as a form of optimized generation?


To achieve these goals, rule-based geometry is generated through the embedding of parameters which respond to optimization criteria later in the design process. A meaningful geometry, which is more efficient, effective and, corresponds with social, economic and environmental constraints, is more significant than undisciplined expressive forms. The value of contemporary design has been profoundly depreciated by the fact that nothing can be surprising any more due to the quick spread of news through media today. Meaningful geometry does not mean universal geometry based on the same principles. On the contrary, it should embrace novelty rooted to a specific site context to harmonise the battle between globalization and localization.


• Problem 3: Experience-based design practice



In conventional architectural practice, design decisions are made based on experiences due to the lack of constant analytic feedback data of the proposed forms and spaces, for example, thermal performance, daylight penetration, people flow, bills of quantity, structural efficiency, and other criteria. Instead of using analytic tools to optimise pre-defined geometries, can analysis-based technologies and techniques contribute to the generative geometric processes?


Conclusion


When new trends emerge, relentless resistance from the camps of well established discourses also tend to react, to safeguard their traditional values. Looking back to human history, technological developments have always brought changes. In many regions, such as China, there are still many who are sceptical that computational design will soon pass as though it were a fashion. The current economic downturn has diminished the proliferation of complex forms generated by computational tools, and the currency of new computational forms faces its biggest challenge from these economic struggles, alongside concerns for greater sustainability.


Synthetic computational design extends the application of computation to architectural industry from focusing on creating innovative forms and spaces, to engaging with the entire process of an architectural project in order to make appropriate design decisions. Architectural industry is now driven by research within its own discipline. Synthetic computational design proposes comprehensive solutions to functions, fitness, innovation, efficiency and aesthetics through the embedding of digital design systems. We should shift our intentions from creating expressive forms, towards an immersion of computational apparatus into every aspect of architectural projects. Creativity does not need to be embodied only through visible forms, but can also be manifested in other less visible actions and procedures of architectural production.






Notes


i Philip Steadman, What Remains of the Analogy? The History and Science of the Artificial, The Evolution of Designs: Biological analogy in architecture and the applied arts, 1979


ii Peter McCleary, A New Concept of Technology, 1988


ii Manuel Castells, Space of Flows, Space of Places: Materials for a Theory of Urbanism in the Information Age, 2004


从标新立异到总体运算化设计

英文：高岩

中文：朵宁

2010/5/17

“设计师发现他自己面临一个特定的历史处境中，某些具体问题的重要性在减弱。他针对这种情景做出特定的设计决策反应，然而这种反应在改变待解决设计问题的同时，又制造出新的问题。 ”

在二十一世纪，我们不断引入新的思考和行为方法，这些方法来源于新科技的发展。技术可以被理解成是各方面技术领域理论的集合。在设计和生产领域同时投入应用的数字化工具对于两个方面都造成了巨大的冲击。参数化设计(parametric design)，或者进一步说，运算化设计（computational design ），作为可持续发展设计的途径之一，展现出新的希望。有意思的是，这两个领域在近日才开始彼此对话。第一代数字技术的建筑师，例如Greg Lynn，Frank Gehry和Zaha Hadid，他们开拓了数码设计的全新领域；接下来的第二代，如Ocean D，Nox，Future System，FOA， Reiser-Umemoto和Ali Rahim等，则从建筑设计的各个领域探索数字化工具的应用；而新生一代难以计数的年轻设计师们，在大规模的应用这些工具的同时，也催生出更切实际的新设计议题。

第一代和第二代数字化设计师们已经证明了新技术的应用不再是海市蜃楼，直接一点的说法是：只要想得到，就能盖出来。所以面对新一波设计师的问题就是：当我们能够盖任何东西时，我们应该如何设计？如果先驱们已经投身于创新，那么当前摆在建筑师面前的问题就是如何把这个创新转化为总体性的设计解决策略。

什么是总体运算化设计？

为了理解什么是总体运算化设计，我们先来考量什么样的运算化设计不能被划入总体设计的范畴。根本上来说非总体类运算化设计偏重设计过程的某个环节，比如新形式，新类别，新原型，相对于总体运算化设计的多重层面，基本上是单一层级和局部剖面。这种设计方法把建筑项目的复杂网络体系分解为片段。总体运算化设计在数字平台的基础上涵盖更为广泛的方面：不光是形式和空间，还有设计信息的沟通和互动；不光是形式生成和自动化执行，还有形体优化和设计评测；不光是数量和质量控制，还有风险管理和造价控制；不光是美学，还有效率；不光是耐久度，还有灵活度；不光是创新，还有适用。总体运算化设计不是妄图，也不可能穷尽一个项目的每个方面，关键在于初期筹划项目的复杂总体，然后运用数字化工具切入制高点，以此来连接项目的不同层面。一个总体式的态度是至关重要的。

但是，光有一个总体统筹的考虑是不够的。各种交叉学科的知识应当通过数字化的运算工具在同一个可量化的平台上整合起来，在这个过程中各种输入被量化和转译成能被提供输入的更各方所共享的数据。设计决策应当建立在对这些不同途径输入的总体考量平衡当中，而不是片段化的武断决定。

第三点，总体运算化设计应当考量社会和环境因素的影响。这种考量有多重意义。环境因素比较容易理解，有一系列已经比较完善的物理环境评估系统，比如LEED，BREEME 等等。这方面本文就不展开讨论了。笔者更愿意关注社会方面的设计考量。一个总体式的设计不应当完全建立在设计者强烈的自我意识上，它应当是一个对于错综复杂的文脉的反应，在这种对话的基础上更新原有的文脉。建筑形式很难逃脱一个文化和场地环境的隐喻表达，极端的情况就是象征性。“在这个一体化的世界中，信息的沟通反而变得危机四伏，因而重新建立象征意义成为最为基本的要务。。。建筑，各种的建筑，应当在这个全球化浪潮中立足于象征意义，在流动的空间中确保归属感。 ” 象征性建筑，或者标志性建筑，通过其独特的外表来定义地方性的新文脉，它在一个同质化的城市环境中起到一个重要作用。部分的评论和批评建立在这种标志化的建筑体型上，但是这种批评应该更多地针对建筑的公共空间部分和大城市空间流动之间的脱钩上，而不是形体本身。 建筑形态作为文化表达的一个重要方面，在这个信息社会变得越来越重要。运算化工具和设计为建筑师提供了非常有利的工具，用来跨越不同风格和其背后的主观创造惯性，以此来为真正的创新提供可能。这些工具同时把设计过程从肤浅的拿来主义形式模拟提升到一个全新的境界，即对形式的逻辑得到进一步的理解，因此建筑知识和其象征意义得到紧密的整合。自然生成的形式来源于设计者的特定设计意图和相关学科框架限制的沟通，设计者需要在这种链接的过程中学习如何运用工具结合，并获得针对最初意图的设计信息反馈。一言以蔽之，总体运算化设计会创造出更有意义的标志性建筑。

为什么采用总体式设计？

科学中的简化还原论在上个世纪占据统治地位，这种方法论相信通过研究各个局部，就可以完全理解这些局部的总和——系统总体。很明显，在实验新技术的时候，我们只能在某一时刻解决某一具体问题。实验性设计通常不能做到整体性，因为其本身要求注重在复杂的项目中只能关注某一方面，以此来获得局部的创新突破。但是，实验性建筑设计推动总体设计不可或缺的一个篇章，因为明白分化的个体，是把握整体的前提。

今天我们生活在一个互相链接的世界中。在这些链接中，联系的多重性和互动的复杂性比起个体本身更加具有意义。因此分化一个建筑项目的各个部分，希望借此来理解这些局部组成的整体变得不再可能；而在一个项目提议中穷尽有关项目文脉的所有方面，也是不够现实的。 但是我们可以通过不同学科的交叉，不同知识的撞击，来整合和最大化设计的整体性，而这种整合在数字化平台上变得更加有效率。这种态度的一个类比就是东方医学和西方医学的对比：西医注重表象的痛楚感，而中医研究导致疼痛的人体经脉循环系统。西医见效快，但是副作用强而且不能够治本；而中医疗效缓慢，但是副作用小且通常作用于疾病的根源。很明显，中医的方向是长效性的，因而更具有可持续性。总体式设计也是这个道理：对于建筑项目的理解是一个复杂网络，对于具体问题的处理是牵一发动全身。

直到现在，还有很多讨论，认为运算化设计只是针对造型夸张且造价昂贵的那些方案。这些批评通过表面上的不规则的外形来评判设计结果，但无法认识到运算化设计的过程意义，思考方法，建造手段和质量控制，而所有这些环节都是通过运算化工具的应用得以链接和实现的。运算化设计已经从沉迷于表面形式的游戏发展为对于建筑的整体设计和使用跨度中的智能、效率和自动化的考量。一针一线的紧缝密补，整体式设计在逐渐渗透到建筑工业的各个方面，早已经超越了表面的形式和空间游戏。

总体运算化设计和可持续性发展

没有人怀疑可持续性发展设计是二十一世纪的主题曲。创造一个可持续发展的人居环境，不仅仅满足人类当下的需求，还需要考虑到我们的后代。这种理念正在变得如此深入人心，不久的将来不会有人再认为这种设计是额外的考量。我们认为运算化设计也必然走上这个方向。运算化设计的核心是优化，这与可持续发展设计的本质息息相关，例如能源消耗的最小化和建筑效能的最大化。我们应该尽快建立可持续发展设计和运算化设计的桥梁，前者是设计意图，后者是技术与方法。

可持续性发展设计有三个维度：环境可持续性，经济可持续性和社会可持续性。简单说来，环境可持续性通过平衡物质能源消耗和人类活动产生的建筑排放，来关注生态环境的可持续性发展；经济可持续性则围绕着把自然环境作为经济的外部性(Economic Externality)，寻找良性发展的机会；社会可持续性着重于未来能具有同现在等同的社会资源。这三个维度都紧密围绕着建筑行业，因为建筑行业占全球碳排放总量的40%，同时消耗了大部分的原材料。

可持续性发展已经被关注了很多年。在欧洲的很多国家，例如英国，芬兰，挪威和德国等国，可持续发展都被纳入建筑行业规范。在过去的十年中，很多相关产品被研发出来。但是，在大部分可持续发展设计中，还存在三个根本性的问题厄待解决。

问题1：片段化处理

在建筑可持续发展设计中，只有社会，经济和生态因素被统筹考虑，才有可能达到满意的结果，否则可持续性设计容易变成脱离现实的面子工程。在当下也许能够满足气候的要求，但是当地社区的利益和经济动荡的优先性随时可能超越设计初衷。只有总体式的设计态度才能够在三个维度平衡可持续性发展的需要。这意味可持续性设计本身也需要更可持续，或者说前瞻性的考虑。这一方向只能以数字平台来整合各方面设计因素，将其量化为数据，输入运算化模型中总体协调。

问题2：建筑空间和几何形体的内在逻辑缺失

建筑的物质属性决定了很多可持续性设计的考量集中在建筑材料、能源消耗和供给、建造方法等方面。设计实践变成在设计形体之后选择正确的材料，把这些材料安置在合适的框架之内。建筑设计的空间和形式变得与材料脱节，后者业主根据材料供应商的条件解决。设计在工程后期必须要频繁修改，以适应建筑材料的限制。这样的工作流程严重耽搁了设计时间和能量。

在这个信息过剩的时代，新材料不断的涌入建筑市场，建筑师不能再担当起熟悉每种材料的工程大师的责任；相似的道理，完全洞悉不断复杂化和分散化的建造方法也变得日益困难。建筑师最终只能控制受各种参数影响的总体空间与形式。当建筑师开始最初的几何大形草图时，他们有可能加入可持续性发展设计的考量么？相对于一般的从设计生成到优化的过程，有没有可能让优化发生在设计的第一时间，或者优化指导着生成过程？

为了达到这一步，必须要设立规则来引导几何形式的生成，而这些规则应当嵌入回应优化标准的参数，来为后期的深入作为铺垫。一个有意义的几何形式，应当在回应社会，经济和环境限制方面更加有效。这比一个徒具表现力但并无实际意义的体型要更加有意义，因为后者在一个媒体泛滥的社会中即刻会贬值，或者被更引人注目的体型所超越。笔者提倡的有意义的几何形体，不意味着回到现代主义的简单几何形体：在一个全球化和地域化交错拉锯的文脉环境中，有意义的形式意味着创新来源于两者的互动，即通过调动全球资源让设计植根地域文脉。

问题3：经验导向的设计实践

在传统设计中，设计决定建立在设计师的从业经验上，因为针对特定形式和空间，有很多反馈参数是缺乏精准的量化数据的，比如热效应，日照分析，人流分析，材料清单，结构优化等等。传统的方法是通过运用模拟模型来优化已经确定的几何形体，那么有没有可能让这些优化过程作为生成几何形体的机制呢？

结论

当新的思潮涌现之时，总会在已经建立的主流话语中搅起轩然大波。如果我们回溯人类历史，技术发展驱动的变革总是不可阻挡。在很多地区，比如中国，仍然存在大量对于运算化设计会像潮流一样转瞬即逝的怀疑。当前的全球金融危机也在另一方面冲击着运算化工具生成复杂性体的局限性。经济的不景气和对于环境的忧虑，同时刺激了开拓运算化设计新领域的努力。

总体运算化设计扩展了运算化设计的建筑实践，从关注形式，转向整合设计资源，目标是做出更加可信赖的设计决定。这也许是第一次，建筑行业在其自身内部寻求创新的动力。总体运算化设计提倡对于功能、适用性、创新、效率和美学的综合性解决方案，这一过程以数字工具为基础。我们应当放开眼界，把重心从对于形式的关注转移到运算化工具在建筑项目各个环节的综合运用当中。创造力可以被表面的几何形态所表达，也可以被不可见的建筑内部机制所体现。