In 1985, the world’s best human chess player of the time – Garry Kasparov – simultaneously played against thirty-two computers. He comprehensively beat all of them. In 1996, he narrowly beat the world’s most advanced chess-playing computer. In 1997 he was comprehensively beaten by the same machine. In 2016, a computer repeated the feat in the more difficult and ancient game of Go. In 2012, as a fitting tribute to Artificial Intelligence pioneer, Alan Turing, the London Symphony Orchestra played music that was composed entirely by a machine – Iamus. The composition was widely applauded for its expressiveness and sufficiently intrigued human musicians.
Digital technologies – computers and computer-controlled machines – have pervaded all aspects of life, delivering sustained and accelerated rates of societal and economic evolution. Yet, in Architecture, such an embrace of digital technologies and attendant intellectual disposition is not widely accepted. [1] Whilst increasing numbers of progressive firms and research institutions are forging rapidly forward in such a technological upgrade, the larger populace of architects and the architecture produced thereof is decidedly averse to it. This unfortunate and hopefully temporary resistance notwithstanding, digital technologies will incontrovertibly be one of the key drivers of Innovation of our discipline and consequently the built environment in the 21st century.
Artificial Intelligence and Intelligence Augmentation
Closer inspection of the previous examples of machines bettering humans at innately human tasks, reveals some nuances. Following the cataclysmic event in 1997, the quality of human chess players world-wide improved dramatically by incorporating software in their training. [2] Additionally, and interestingly, human-machine combinations routinely outperform super-computers and super-humans. This can also be observed in the more recent development of Go players, [3] robotic musicians accompanying human musicians, [4] in post-natural disasters rescue operations, etc. [5] Computers and Robots are contrary to popular belief, making humans better. Rapidly better. This lends credence to the under-rated but seminal hypothesis of human-machine symbiosis by: [6] In the long term future it seems entirely plausible that an Artificial Intelligence will dominate and more pragmatically, in the near future there is an exhilaratingly rich period of symbiotic progress to be worked and capitalised on. In other words, we are in the Intelligence Augmentation phase of human evolution. Architecture should not and cannot afford to be marginal to this. This article will, outline the two critical endeavours to exploit this innovation potential – A framework of Architectural Knowledge and Collaborative design practice. The article will also illustrate the same with exemplar projects from Zaha Hadid Architects. (Suggested: Two critical endeavours to exploit this innovation potential are methodically condensed by a framework of Architectural Knowledge and Collaborative design practice).
Architectural Knowledge
What could be the nature of an Architectural Knowledge that provides the foundations for Architectural production in the digital age? Inspecting the concerns of such a Knowledge base, one would discern two major divisions: one aspect concerning itself with the Technologies of Design and Construction and the other with the Conception of Design. The former concerns itself with operational knowledge of computers and machines, material behaviours, structural systems, use of specific software, programming etc. The latter includes knowledge related to spatial organisation, Styles of design, socio-cultural implications, and theoretical schools of thought, etc. Witt poetically traces back such divisions, at least back to the famously contrary positions of the two protagonists of the Italian renaissance – Filippo Brunelleschi and Leon Batista Alberti.[7] [8] He suggests, implicitly inclined towards Brunelleschi that current digital age could learn substantially from the 19th century efforts in systemic generation of architectural knowledge of the first kind - abstraction of mathematical knowledge into drawing instruments for specific types of complex geometry, manuals of construction for their physical realisation, etc. Schumacher, inclined towards Alberti’s efforts, argues for a similar effort. In this sense, the incorporation of computational tools and scientific methods in the conception of design is widely discussed as well – specifically in the study of human perception of spatial features and the subsequent production of Architectural meaning. [9] In other words, a computational understanding of semiosis that can then be used to generate spatial constructs that enhance such a process. This in turn helps humans to navigate the spaces harmoniously, complement and augment human activity within the space, etc. Thus a fundamental necessity for sustained innovation is the development of a computational basis for both aspects of Architectural Knowledge – Technological and Conceptual – and further their unification into a framework. There have been several, episodic and partial attempts at such frameworks that flourished in the 1960s and 70s. [10] The necessity now, is for unhindered, devoted pursuit and expansion of the same. This is imperative for architecture to be able to deliver similar accelerated rates of evolution as some of the other aspects of human evolution previously mentioned.
Collaborative and Co-authored Design
The nature of relationship between architects, engineers and contractor-builders in post-Renaissance history, has fluidly oscillated between being unified to being distinct and domain specialised.[11] These role changes between architects and engineers are fascinating. Initially the two professions were indistinguishable on the basis of skill, but more by building task (civil buildings by architects, bridges by engineers for example). By the 20th century, the roles had emerged to account for division of labour on the same building project - Architect Sauvestre inflecting Engineer Gustave Eiffel’s Tower, or Architect Utzon’s Sydney Opera House being physically realised by engineer Ove Arup. In the present time, the increased use of digital means in the design of the spatial, geometric aspects along with the structural and construction aspects of building, presents an opportunity for increased collaboration and co-authorship of design – i.e. a relationship situated between the domain general, ambiguous distinction of the 18th century and the domain-specialised, hard distinctions of the 20th century. The use of computers provides a unifying platform between various disciplines, especially in the early generative stages of design. Thus the computational medium allows for the (equal) participation of the not only the traditional stakeholders of design process – architects, engineers and builders, but also other sciences that operate using the medium – mathematicians, biologists, sociologists, and so on. Thus, the 19th century architect, Antonio Gaudi could draw inspiration from the biological ideas and drawings of Ernst Haeckel, or develop an artistic repertoire influenced by the formal appearance of new mathematics of the time. [12] Contemporary computational designers on the other hand, can use the very biological models that generate our physiology to produce geometry of architecture. [13] [14] They could, in equal part utilize the code of complex mathematics to generate the structural systems as in the Beijing water-cube stadium.[15] However, legally and in the final execution of the projects, hard distinctions are productive and necessary. Thus, digital technologies can allow for fluid transition from a co-authoring early stages to a collaborative, specialised later stages of design and execution.
Exemplar project – Mathematics: The Winton Gallery at the Science Museum, London
The Analytical Engine weaves algebraic patterns, just as the Jacquard loom weaves flowers and leaves. - Ada King LovelaceThe design atelier of Zaha Hadid, founded in 1979, was an early pioneer in and adopter of both these key necessities of Innovation – systemic knowledge generation and collaborative design. In line with those efforts, the Computation and Design research group (ZHCODE) of the company was an effort endeavour initiated in 2007, in line with the preceding pioneering efforts of the company. The explicit aim for the research and development efforts of the group was to harness the opportunities latent in the inter-disciplinary collaboration of computationally literate architects, engineers and emerging digital manufacturing methods. The atelier has now grown into a large firm with several seminal built projects in this new paradigm of Parametricist architecture. It would only be fitting to describe one of the latest projects to embrace the ethos – the design for the gallery for Mathematics and Computation, at the Science Museum in London, completed in December 2016. [16] The project is a testament to the aforementioned critical aspects of innovation, collaborative design processes and the fluid exchange of means, methods and models across disciplines.
Conception of Design
Central to any gallery is the curatorial vision and the objects themselves. The architecture augments this vision, spatially supplements the narrative and amplifies the assimilation of the information presented. It is therefore natural to make the objects and the narrative as the motivating driver for the spatial organisation of the gallery. Additionally, if the objects change, the spatial organisation has to accommodate. The approach to this was data-driven. The first step was to tabulate the data - the data of the hundred odd objects, their eighty odd showcases, and their relation to their principal storyline as also the remaining twenty-five storylines, their position within the six categories, dimensional information, sensitivity to light, requirements of preservation etc. Next was to format the data to enable consumption by a data-processing algorithm. A bespoke algorithm then processed the information and laid out the objects to negotiate the often disparate requirements – curatorial vision, object dimensions, ease of navigation, available space, access and circulation requirements, construction costs etc. This enabled the spatial layout of the gallery to be changed easily, if the objects, stories or any another aspect of the curatorial vision were to change. This is often the case to accommodate several vagaries and multitude of stake-holders involved in the commissioning, design, execution and maintenance of a permanent exhibition. Additionally, such a process enables easy measurement of critical performance criteria of the proposed layouts. Apart from the functional metrics such as structural and material feasibility, we were principally concerned with the user-experience of the space. The visual field of an average visitor across several possible access routes were routinely studied and the spatial layouts adjusted accordingly. Primary user navigation and storyline distribution is naturally emphasized using spatial and easy-to-register aspects such as curvature, fluid and interrupted visual field etc. This obviates the need for way-finding signage. This is further accentuated by resonance in several other ancillary features such as the lighting and floor tile layout, colour scheme, height distribution of the showcases etc. All the major features of the space thus become inter-correlated and cohesive with the human navigation and occupation of the spaces.Technology of Design and Construction
Two specific features – the central fabric structures and the bespoke seating design – of the gallery are worth mentioning in the context of historic knowledge of design and construction. These also highlight the need for Innovative Design to follow a research programme, [17] as opposed to ad-hoc solutions to design tasks. Imre Lakatos, a philosopher of mathematics and science, used the word – research programme – both in pragmatic terms of cultivating experience and also the philosophical sense of maintaining a set of core-beliefs (about design in this case).
Fabric Structures
The geometry and materialisation of these central organising features of the gallery are a result of both practical transfer of knowledge across disciplines and also a lineage fabric structures – so called minimal surfaces – that the office has undertaken in the past. [18] Their computational generation – a so-called form-finding process – usually employs one of two popular methods – the Force density method [19] and the Dynamic relaxation method. [20] These seminal methods have been made more accessible to architects and engineers alike by research institutions like Block Research Group [21] and University of Bath. [22] [23] their architectural materialisation as stretched cable and fabric forms has been studied by several architectural and engineering firms including ours. Prominent prior examples include the seminal Munich stadium by Frei Otto, and the temporary Serpentine Pavilion (London), the Magazine restaurant (London), the interactive Parametric Space installation (Copenhagen) by Zaha Hadid Architects etc. Thus the latest manifestation of such structures in the Mathematics gallery is a result of a long history of prior experience and historically assimilated and transferred research.
Wire-cut Concrete Benches
The gallery has several moments of ‘pause’ including fourteen benches – designed as cast, ultra-high performance concrete benches. The shape and physical production of these also owes its development to a long lineage of research in the mathematics, engineering and materialisation of a certain class of surfaces called Ruled Surfaces. Mathematically these surfaces have been known for centuries but it’s in depth study gained traction after the invention of calculus and is widely credited to French Mathematician Gaspard Monge. [24] As mentioned in the introduction, such in-depth mathematical knowledge was abstracted and captured as drawing machines and construction manuals during the 19th century. [25] These inventions in turn, made them widely accessible and their materialisation in stone and timber significantly more feasible. [26] These were very prominent and widely used in the 19th century masonry and timber structures [27] - perhaps most famously by the Spanish Architect Antonio Gaudi, in his church for the Sagrada Familia, Barcelona. The benches on the gallery inherit this mathematical, physical and material history, and employs it in a contemporary setting including a collaboration with state-of-the-art robotic company specialising in hot-wire-cutting of foam [28] to produce the moulds for the cast concrete.
Collaborative Design
As mentioned previously, the computational methods employed to generate the shapes and spaces of the gallery, especially in early stages of design, were a result of a fluid exchange of means, methods and models across disciplines. For instance, the simulation of the air-flow around the key figuring object of the Hadley Page airplane, has a lineage in the physics of Fluids dynamics going back to Gabriel Navier and Claude Stokes in the 1840s. Further they were made accessible and amenable for use in early, interactive stages of design by sustained research in computational fluid dynamics by the likes of Jos Stam, [29] Ron Fedkiw [30] and others from the computer animation and graphics industry. Thus we were able to utilise the actual models and code as opposed to merely drawing inspiration from the formal appearance of fluid-flows. Similar influences on the central fabric structures have already been mentioned. Such inter-disciplinary osmosis in the early stages has now transmuted into more clearly defined roles – architects, engineers, contractors – in the later stages of the project. Industry standard Building Information Modelling (BIM) and similar digital technologies are enabling a well-coordinated execution of the project.
Design Intelligence
Cumulative Progress
As exemplified above, the benefits of following a cumulative research program are thus two-fold. It may be noted here that, as trivially obvious as both aspects might seem, it is far from de facto in current architectural practice. On the one hand, aligning practice-embedded research with established research trajectories allows for practitioners to focus their efforts on the social implications of the built environment, which has, despite its importance and impact, hitherto received rather scant attention from designers. [31] Early evidence of this can be discerned in the design of the mathematics gallery. Additionally, this alignment is also mutually beneficial to the researchers in that their work can be motivated by and tested against its application in the field. These two aspects thus motivate all of the various research strands of ZHCODE and their gestation and development trajectories. Apart from research into structural and ruled surfaces that has borne fruit in the mathematics gallery already, research in curved origami is also at the precipice of being similarly impactful.
Intelligence Augmentation
Parametricist or Generative design has the potential to overcome the mass-produced, homogenous, and disorienting sterility of 20th century architecture. It has the potential to re-associate with historic practise, and amplifying assimilated knowledge. It has the potential to heighten the Inference potential of spaces – of enabling meaningful occupation and navigation of spaces by humans. To fulfil this potential for rapid evolution of our discipline and upgrade of our built environment, it is imperative that designers and other stake-holders of architecture, invest in it – invest in digital technologies not just digital means of producing known tropes, invest in making design processes amenable for the use of computers, invest in making materialisation of architecture amenable to the use of robots. Digitisation of architecture and Intelligence augmentation of designers is a necessary and imperative path to a superior design intelligence.
