Flexible Design and Scripting: a computational tale in the Digital Era [part 1]
In 1947, the Italian architect Luigi MORETTI coined the expression: “parametric architecture”. An established post-war Italian architect, at the end of the 1940s, MORETTI had the intuition to integrate mathematics and information technology into the traditional approach of architectural design. In 1957 he founded, in collaboration with the mathematician Bruno DE FINETTI, the Mathematical and Operational Research Institute for Urban Planning (IRMOU). Through the adoption of state-of-the-art computer equipment (610 IBM — one of the first personal computers) and the collaboration with researchers from multiple scientific fields (mathematics, physics, psychology, electronics, biology, philosophy, economics, sociology), he defined a research path of the architectural language that did not reside in the dichotomous “form-function” relationship but, rather, in the “process that leads to its definition”. The designer would therefore have the acumen and sensitivity in recognizing and interacting with all those parameters necessary for the materialization of the architectural project.
In 1960, at the XII Architecture Triennale in Milan, it was possible to admire the incredible result achieved by the IRMOU research group by showing graphs and plaster models of the football stadium [1] made according to the new design vision resulting from the mathematical and computer interdependence of material and immaterial parameters (around 19 params included the ticket cost).
The scientific literature recognizes in Moretti, and in his works, the paternity of using the parametric terms applied to design and architecture even if antecedent studies dating back to around 1837 report the use of “parametric language” as a means of the physical description of the model three-dimensional crystal attributed to the scholar James DANA [2] — paper On the Drawing of Figures of Crystals. Once the procedural dependence on material or immaterial parameters has been introduced, the modern vision of design can be conveyed according to the increasingly high demand for precision and verification of project data and the spasmodic need to receive recursive feedback from the client or from other members of the teamwork in real-time. All this makes it necessary to adopt a “flexible” approach to design. In this discussion, flexible design means an operational process aimed at identifying and discretizing the criticalities of the topic addressed (problem-necessity) by formulating relationships between parameters (geometric, quantity, function, environmental, etc.) that contribute to the definition of a family of solutions that respond to a predefined performance-based verification system. As the researcher Francesco DE LUCA reports, referring to parametric design, in particular to the associative technique [3]:
the purpose of diagrammatic design using Script and Associative Modeling is not to find the best shape or the only solution for a given design problem, but to generate variations and configurations providing the designer with a framework of possibilities
That is, a sort of bottom-up design as it focuses on the relationships between individual elements that trigger a mechanism for the dynamic generation of possible configurations. The formulation of relationships between parameters is possible thanks to the scalability of the process provided by the IT tools available today and the diffusion of design software capable of guiding the designer in the “rigorous construction” of his own model. Thanks to the explicit introduction of programming languages integrated into the same development environment, it is possible to write codes (scripting) in parallel with traditional design activities.
Without going over the history of applications and Software Houses in detail, it is useful to note that some players, of the calibre of Autodesk, had already adopted this policy of integrating programmable parametric modules, for example, Autocad (1982) already existed with its AutoLISP language; Pro/Engineer and Dassault Systèmes with CATIA v4 (1993). Without taking into account that in the early 2000s BIM modellers such as Revit and Archicad began to consolidate.
Scripting is a practice that allows you to interact synergistically with the chosen development environment by writing “rules” (code strings) read in sequence by the compiler (module inside the software that automatically implements and verifies the new rules/functions). The set of rules translates into executable “instructions”, that is, into a design algorithm.
In this way, whatever the result of the modelling performed by the designer, it will always be possible to retrace and modify the individual steps that led to a certain result. The integration of the computer language will always be characterized by an input parameter — algorithm — output parameter. Below is the definition of the algorithm provided by Ipek Gursel DINO [4], a computational design scholar:
An algorithm is a finite set of instructions that aim to fulfill a clearly defined purpose in a finite number of steps. An algorithm takes one value or a set of values as input, executes a series of computational steps that transform the input, and finally produces one value or a set of values as output
Parametric design is ultimate to be understood as an integral part of algorithmic design. There are multiple platforms used by professionals that allow you to perform programming from scratch capable of enhancing the basic capabilities, and functions, offered by the development environment used. Tools that intrinsically involve a fervent programming and design activity since their use is constantly based on conditions and evaluations — paraphrasing the expression of Peter EISENMAN [5] who considers them as the “everything and nothing of the project […] a white writing”.
It is known that the technology transfer that took place in modelling software came from the film and mechanical engineering industry [6]. So we find the MaxScript language in Autodesk 3D Studio Max, the Maya Embedded Language (MEL) in Autodesk Maya and, in the last 6 years, the modelling environment of McNeel Rhinoceros with its Visual platform is increasingly taking on a leading role. Visual Programming Language [7](VPL) Grasshopper, but this is not the only computational design platform on the market. For the record, the same BIM platforms are equipping themselves with proprietary environments delegated to parametric/algorithmic design for example, in the case of Revit, you can have Autodesk Dynamo (2011 in beta) capable of managing native BIM objects; Vectorworks with Marionette (from 2015 version); while GRAPHISOFT, with Archicad, has undertaken a close collaboration with Grasshopper since release 18 (2014) through a special addons, LiveConnection AC-GH, boasting the BIG [8] design studio as a prominent tester. Another valuable VPL project since 2014 is Sverchok [9].
This information landscape is useful in making two very important observations that have an “instrumental” and properly “cultural” value that is reflected in the professional life of the architect and designer, but also in the world of research in architectural technology.
In the early 2000s, there were many companies intent on introducing programmable parametric modules within their development environments, Bentley Systems [10] was one of them. In 1985 he proposed his first release of the Microstation CAD software, a platform for 2D/3D modelling of products marketable by Bentley itself. The direct competitor of Autocad, in 2001 in version 8.0, was already able to communicate with the “dwg” CAD data exchange format. The introduction of the GenerativeComponnents (GC) module developed by John NASTASI in 2003 (Institute of Technology of New York) marked an important step both from a commercial and academic point of view as its development, and related beta testing phase was often entrusted to the SmartGeometry [111]. research group, one of the first non-profit associations that established for the first time a solid commercial and cultural axis between professionals and the world of universities. Below is one of the main focuses proposed by the association:
Computation and the use of computer as an intelligent design aid provide the focus for the SmartGeometry Group. […] Geometry is one of many systems which are amenable to modeling. Architectural design as a process should, we maintain, exploit the new potential available in computing. To the new generations of architects mathematics and algorithms are becoming as natural as pen and pencil.
The activities of the SmartGeometry Group promote the emergence of a new generation of digital designers and craftsmen, who are able to exploit the combination of digital and physical media. The group’s interests range from parametric design and scripting to digital manufacturing.
As the SmartGeometry Group has developed it is moving its focus from tools and techniques to a forum where the new critical language of emerging architecture can be formed
In 2005, the software ecosystem proposed by Bentley Systems was already widely introduced in the London AEC industry. In 2007 it was commercialized on a large scale. This historical sequence made up of commercialization and disclosures in the academic field (thanks also to the presence of professor Robert AISH — currently a consultant at Autodesk — one of the founding fathers of Bentley Building Modeling and GC), represents the real paradigm shift in modelling that moves inexorably attention is paid to the definition of procedures [12]. The probability that the fervent English climate fueled by Bentley’s work has contributed greatly to the role of leader assumed by Great Britain in the battle to adopt the “BIM philosophy”, is very high and it is a key element in shaping the international balance in the political and professional framework of the construction industry.
(below) Bentley Systems GenerativeComponents user interface
However, the work done by Bentley, and the commercial policy of the time, was not exempt from being the protagonist of a typical episode of missed “innovative vision”, history is full of these events.
PART 2 will be posted next week, stay tuned!
You can find the 🇮🇹 version on my website here
References
to cite this article please use this:
Ambrosini, L., (2018), Data, Digital & Design — Production of the digital project and decision-making processes: “flexible” design in the Era of Scripting and Building Information Modeling as a new technological paradigm, PhD thesis in Architecture. XXXI ciclo, DiARC, Università di Napoli “Federico II”. DOI: 10.13140/RG.2.2.27158.29769
[1] Cfr. Bucci, F. e Mulazzani, M., 2000. Luigi Moretti: opere e scritti [Ristampa], Milano: Electa.
[2] Cfr. Davis, D., 2013. Modelled on Software Engineering: Flexible Parametric Models in the Practice of Architecture. PhD Thesis, p.19.
[3] F. De Luca, Diagramma and Script, in A. Saggio, 2011, Architettura & Information Technology, m.e. architectural book & review, Roma, pp. 82–83.
[4] Dino, I.G., 2012. Creative design exploration by parametric generative systems in architecture. Metu Journal of the Faculty of Architecture, 29(1), pp.207–224.
[5] Peter Eisenman, 2004, Giuseppe Terragni: trasformazioni, scomposizioni, critiche, Quodlibet, Macerata.
[6] A list of programming languages that gravitate around the world of design: Adobe ActionScript, C, C #, C ++, Generative Components script, Html, Iron Python, Java, JavaScript, LUA, Mathematica, MATLAB, MaxScript, Maya (Maya Embedded Language ( MEL) and Python), Objective-C, Perl, PHP, Processing (Java), Python, RhinoScript (VB, Grasshopper (VB), Python), Rhl, VB, VBA.
[7] It is a language that allows computer programming by manipulating graphic objects bypassing formal syntax. In most VPLs, the syntactic scheme consists of “boxes” and “connectors”, the former carry code (instructions), while the latter weaves the classic flow diagram structure. Often the produced source codes (algorithms) are nicknamed “Spaghetti Code”, due to the massive use of branches and connections.
[8] Copenhagen-based architecture firm founded in 2006 — Bjarke Ingels Group (BIG) website: www.big.dk
[9] Sverchok is a parametric CAD tool, built to help generate complex 3d shapes using a node system to control the flow of math and geometry. It is ideally suited to Architects and Designers, but anyone with high school Math and Trigonometry will be able to produce results that are impossible to achieve unless you know text-based programming languages such as Python or C sharp. It is an add-on for Blender. Sverchok consists of two addons — Sverchok itself and Sverchok extra tools from Illya Portnov. Both are on GitHub.
[10] Bentley is a global leader dedicated to providing engineers, architects, geospatial professionals, constructors, and owner-operators with comprehensive software solutions for advancing infrastructure. Founded in 1984, Bentley has more than 3,000 colleagues in over 50 countries, more than $600 million in annual revenues, and since 2011 has invested more than $1 billion in research, development, and acquisitions. Sito: www.bentley.com
[11] The SmartGeometry research group (non-profit organization) was founded in 2001 by Lars Hesselgren of KPF, Hugh Whitehead of Foster + Partners and J. Parrish of Arup Sport. The group’s goal is to trace a significant link between practice, research and academia by promoting the dissemination of computational culture and computer-assisted intelligent design. Main partners: Foster + Partners, KPF, Grimshaw, Arup, Buro Happold; University: Architectural Association, MIT, Delft Technical University, University of Bath.
[12] From a geometric point of view, what matters is the set of numerical, algebraic relationships, etc. that are established between the elements. To prove the truth, the file produced by GenerativeComponents (.gct) is called “transaction” and cannot be read by a CAD, but its result can.
