advanced parametric tools for sustainability in architecture

cesare griffa tinkering desk 2012

Below, an abstract of the Research Brief Proposal submitted to the Fulbright Commission to develop my research on smart creatures. Riding  through the topics of parametric design, rapid prototyping,  electronic hacking, biological hacking, with the final destination of setting up a solid background for innovative sustainable building technology design.


Parametric design methods applied to the production of sustainable architecture

Project Statement

Based on the extensive use of advanced parametric design tools, rapid prototyping techniques, electronic hacks, and chemical labs, this research is committed to investigate strategies for developing a parametric design method addressed to sustainable architecture. In particular, it’s focusing on conceiving architectural organisms as populations of intelligent components challenging formalistic, functional, and fitness issues. On the one hand this research brings to the analysis of large scale population methods based on complex parametric organisms. On the other hand, the small scale focus is on the individual single intelligent component to be used to populate larger scale organisms.

Advanced parametric tools

Parametric Design Tools make possible to define the forms of architecture through a series of complex mathematical algorithms. Then, the issue of form generation switches from a sculptural problem into a mathematical problem. The algorithms that describe forms are generally addressed as “codes”. The interpretation of design as a coding activity opens a new perspective for architects consisting in the possibility of describing their architectural creatures through the genetic mathematical algorithm that lies beyond forms.

Working on codes has always been the specificity of computer programmers. However, the releases of design software like Rhinoceros or Maya are moving towards the introduction of graphic interfaces allowing to work in a parametric environment also without having specific coding knowledge. Coding is

no longer a job for programming experts only.

In recent years, the evolution of design software is bringing to a similar situation to the one that happened over the ‘90s with the WYSIWYG (What you see is what you get) web-software that opened up the possibility to generate HTML (Hyper Text Markup Language) pages also to people non familiar with HTML language.

The most powerful parametric design tool at the state of the art is probably Grasshopper, a plug-in for Rhinoceros NURBS modeling for Windows. Grasshopper graphic interface allows the user to do logical operations that affect 3d geometries in the modeling space of the software. The built of a logical system is named “Definition” and corresponds somehow to the DNA of the visualized form. Logical operations are based on a INPUT OUTPUT system, in which it’s possible to introduce data variations in the INPUT (data) triggering variations in the OUTPUT (final form) without changing the series of logical operations (Definition).

To make an example, we can imagine to write a parametric Definition that creates a series of circular shapes. We need a INPUT to begin the process, that could be for instance a point list of center points and a series of radius. Once informed with the INPUT the Definition will create as an OUPUT a series of circles. It’s pretty simple to imagine that to every changing in the INOUT list will correspond changing in the OUTPUT result, without any variation of the Definition itself.

This design method is very powerful in the creation of series of variable components based on the same geometrical primitives, but customized on very specific local conditions. The process becomes very interesting if the INPUT are environmental data like temperature, light, humidity, solar energy, etc. We can imagine hyper efficient architectures built with intelligent components that behave very specifically on micro-local conditions.

Rapid prototyping techniques

An important problem of such architectures concerns the possibility of building in an economically efficient way design in which all the components are different the one to the other. It’s an approach that moves beyond the classical industrial approach, based on the effort of unifying all building components, towards a direction of digital manufacturing recovering the values of hand crafted products customized on very special needs. The value of the industrial process could be measured by its ability to avoid imperfections in the artifacts, by its ability to produce objects absolutely identical to each other. This system contrasts significantly with the craft: the craftsman quality is measured not in terms of uniformity among the manufactured products and the ability of quantitative production, but is measured as a function of product quality. Often very good craftsmen care products and then take a long time on each piece, and consequently we are not able to produce large numbers. On the contrary, a good industrial process manufactures many products of the same quality. The large quantity is a financial obligation for a good industrial system, whose installation costs are very high: the only way is to divide the cost of amortization by the greatest number of products.

The industrial production systems are characterized by enormous problems of environmental impact mainly because of the large amount of energy and land required for production. Often profit optimization calculations require the use of non-renewable sources in large quantities to produce the necessary energy.

Digital manufacturing, often called rapid prototyping, consists in the use of different cross-articulated software driven manufacturing services that allow a relatively optimized physical production of components with complex variable geometries. These production services are usually called CNC production machines (CNC – computer numerical control). They are based on a technology that interface directly the production machines such as milling or other machine it with computers. These technologies were born in the 70’s, and have recently undergone major processes of optimization and the costs are rapidly dropping down allowing for larger diffusion. It is now rare to find craftsmen who don’t use CNC machines, reinventing old works, which often tended to disappear. The technological artisan, sometimes called “craftsman 2.0”, is a new figure in the world of manufacturing production. Its methods of operation allow to combine advantages of industrial production processes, like high precision, with the benefits of handicraft production, such as the high possibility of customization and product variations.

CNC technology for production of objects is mainly divided into two families: material subtraction and material addition. Material subtraction technology (milling or laser machines) operates by subtracting material through a carving process in larger material blocks. Additive technology is based on 3d printers that add layers of material one after the other to create complex objects.

The use of digital manufacturing techniques is at the beginning of its era, and seems to be nowadays the most efficient system to produce series of customized components, similar among them, but slightly different according to micro-local needs. These components can be designed through the use of parametric design tools, and built with digital manufacturing machines.

Electronic hacks

Electronics can be seen as the nervous systems of manufacturing productions. They run inside industrial objects like cars or computers, and make them sensible to what is going on around them through the use of human interfaces or environment sensors. Manufactured products can become interactive through the use of electronics.

To be interactive, the objects must have a sort of a nervous system capable of perceiving the environment variables (INPUT), go through an elaboration based on a system of artificial intelligence, and provide the answers through various types of actuators (OUTPUT). The state of the art provides interactive designers with a series of small circuit boards, called controllers, that can be connected to sensors (movement, temperature, etc) to collect INPUT DATA, computer programmed to elaborate DATA, and connected to various kind of actuators (Servo motors, fans, LED, etc.) to generate OUTPUT behaviors.

These electronic devices are routinely used by interaction designers to make interactive products and are generally quite expensive. A little big revolution in this world was introduced by Arduino boards, designed and manufactured in Italy by researchers related to Olivetti, and distributed with a commercial open source policy. This process allowed to reduce the prizes in a drastic way, rising a very large community of designers and electronic geeks that eventually could experiment and create interactive systems at low cost.

The numerical control technology, its ease of use andreduced costs, in recent years have contributed to the emergence of a new culture of DIY (Do It Yourself – DIY) in which everybody can potentially imitate the work of the professional craftsmen. This community often works in a new kind of shared space called “Hacker Space” of “Fab Lab”. The FabLabs are small laboratories, equipped with computers and numerical control machines in which objects can be produced. In FabLab typically take place activities of semi-professional manufacture of small series of objects highly customized, and training activities dedicated to students and enthusiasts of modeling. The FabLabs are connected in a global network established at the Center for Bits and Atoms at MIT in Boston, and exist in almost all countries of Europe and the United States and in many countries of Asia and Africa. A FabLab is a place where you are and where you experience the workshop production of unconventional artifacts, which can then be implemented by electronic components to capture a specific intelligence.

The use of digital prototyping technology makes possible to experiment architectural material components equipped with an artificial intelligence software designed using parametric modeling tools. For example, it’s possible to design complex architectural surfaces, like building skins, populated by a multitude of different components with specific intelligence, and test their physical essence through digital prototyping, to highlight problems related both to their manufacture, and their interactive behavior. This low-cost trial is the first step in a path that can go towards larger scale custom productions, to create sustainable intelligent architectures..

Large scale populations

What happens if we consider an architectural surface as a complex system composed of a multitude of intelligent components? In other words, what happens if we consider a skin or a housing complex as an artificial society composed of individuals from a multitude of artificial intelligence? In the human world society, the social sciences are concerned with studying social groups, relationships between them, collective dynamics, political systems, economics, and in general all relational characteristics between individuals and / or groups of individuals. Among all the social sciences, some fields of study of social psychology seem to offer useful insights on how to articulate an argument on artificial systems consisting in a multitude of artificial intelligent individuals.
A useful concept is the theoretical formulation of Kurt Lewin on individual behavior as a function of two variables: personality and environment. The starting point is the assumption that the behavior of an individual in an ideal neutral space would be different from its behavior when immersed in a specific social space, and will change when immersed in a different social space. In the case of human society, the complexity is very large and it is almost impossible to identify all the factors that influence behaviors. In the case of artificial components, it seems easier to argue the case. A first major difference between a loner individual and an individual that is part of a group lies in the capacity / need for interaction with others. The behavior of a single scheme is based on three stages: receiving (INPUT), processing, reaction (OUTPUT). At a time when the individual is to collaborate with others, a part of the INPUT will be the result of OUTPUT generated by others, and even the OUTPUT generated will be used as part of INPUT from others. This dynamic can acquire great complexity and become difficult to control.
The type of system we’re talking about can be compared to natural ecosystems. We consider an ecosystem as a set of individuals of different species living together in an open environment. The interactions of various kinds that are created between all actors, both animate and inanimate, create an equilibrium that may be more or less stable. Ecosystems are generally open, and receive external energy (solar for example) that is used to power various individuals. The balance of the system is of paramount importance. In environmental systems (natural or artificial), it is said that the balance can be positive or negative. It is positive when the matter / energy produced by the system itself is higher than the one consumed. It is negative when this relation is lower. The goal of all artificial systems is to achieve at least a neutral budget: necessary energy is equal to produced energy. In this context, probably the most advanced research are nowadays carried out by the world’s space agencies, whose goal is to recycle 100% of the organic material waste produced within the spacecraft, including the transformation CO2 produced by the astronauts, the autonomous production of food, and to reach energy self-sufficiency by exploiting solar energy.

Creating architectural ecosystems designed with parametric tools, tested with digital manufacturing techniques, implemented by electronic nervous systems seems to be a quite important challenge in order to explore new territories for sustainable architecture.

Intelligent components

Intelligence is generally regarded as a characteristic of beings who oppose their ability to understand, process, decide to the “stupidity” of inanimate objects. Several branches of psychology explore the concept of “intelligence” analysing modes of action and reaction of human individuals, even in terms of their inherited baggage and environmental circumstances, trying to understand how physiological responses are linked to behaviors and environmental variables. The theory of multiple intelligences developed by Gardner in the 80’s identifies a number of specific intelligences attributable to well defined situations. Gardner proposes nine types of intelligence: logical-mathematical, linguistic, spatial, musical, kinesthetic, interpersonal, intrapersonal, naturalist, and existential.

The important aspect of the theory of multiple intelligences is the transition from a paradigm of unitary system in which the scientific measurement of IQ was considered a way to evaluate the intelligence of a real person, to a more relativistic paradigm in which an individual can have a great intelligence only in certain specific areas. For example, assume that an architect can have a good spatial intelligence, and perhaps a more limited linguistic intelligence.
The idea of specific intelligence brings into play the need for cooperation between individuals. The project actions carried out by groups of individuals will most likely be able to cover more areas of intelligence with respect to actions taken by individuals. For this to happen, an effective system of communication between individuals, the availability of every individual to compromise for the good of the joint venture, and a good overall coordination are needed.

There is also another kind of intelligence, studied by the sciences as artificial intelligence, that belongs to the systems of man-made machines. Not all machines are equipped with artificial intelligence. For a machine to be intelligent, it is assumed that it is able to process the “reasoning” in response to INPUT, similar to the human mind. The problem of artificial intelligence seems to be primarily a programming problem.

Programming a computer means to write the instructions that the machine should run. Two aspects are of paramount importance in this operation: the first is tospeak the right language, the second is to set up a good logical structure. Speaking the right language means writing computer instructions that can be processed by the system processor, memory and electronic circuit. Since computers are primarily based on complex systems of electrical circuits that work with an infinite number of on / off switches (1 / 0), the problem of speaking the right language is a primarily math problem: you tell the machine which switches are to activate (1)or deactivated (0), and in what sequence. The machine language can be seen as a communication tool that the teacher (the programmer) uses to teach instructions to-the student (the machine). The programming logic is a general-structure that associates consequences to actions. The best known example of computer logic is the procedure called “if … then … else “. The programmer, in this case, use the machine language to tell the computer what to do if a specific event happens, and if instead that event does not happen.
When working on the single components design, for them to be intelligent, it is mandatory to set up a good logic system, so that every component can elaborate INPUT in an efficient way, in order to give back pertinent and useful information to the general system. To have an efficient sustainable architecture, conceived as a large scale population system, the work has to start from bottom going up: every part of the system has to be efficient.

Sustainable architecture

Sustainability is nowadays one of the main challenges also for architects. The greenhouse effect, CO2 emissions, the depletion within the proximal me 40 years of fossil fuel reserves, the problem of energy supply, pollution issues have become of paramount importance not only in international politics, but also in consciousness of the individuals who populate the planet. The direction is the one of reducing energy consumption and polluting emissions, but the problem seems more complex: how is it possible to deeply reorganize the entire design process from a sustainable approach?

Most trials focused on sustainability pay great attention to the concepts of performance and efficiency. A good process maximizes the result and minimize consumption. This is about energy: maximum OUTPUT with minimum INPUT. One of the cornerstones of the energy savings is to transform the existing energy,-using it instead of fighting energy with other energy. For example, passive systems in architecture try to convey natural flows of air in order to maintain the ideal temperature conditions instead of fighting energetic warm air with energy consuming cool air. Another important issue is the ability to produce energy through the use of renewable sources like sun or wind, instead of using non renewable carbon based resources.

The issue of efficiency of a process or an artifact has to do with his behavior. If the artifact was a living being, some considerations about his behavior at both individual and social level could be done: a responsible behavior would be able to perform efficient actions responding to boundary conditions, and to enhance performative actions from the other individuals of the community. This behavior would be based on the understanding of the needs and requirements of specific conditions, the treatment of possible actions and the evaluation of consequences. In other words, to behave in a responsible way is simply to develop the ability to survive individually and socially through a process of adaptation to the environment. In 1859, Charles Darwin assumed for the first time in human history that individuals belonging to a population evolve over generations to better adapt to their environment. The theory of natural selection is still controversial more than 150 years after its first publication. According to this approach, not only intellectual intelligence but also the physiological capacity of a living organism has the ability to change over time in order to improve the behavior of efficiency and guarantee the survival. Both the mind and the physical are fed by INPUT coming from the outside world. They digest them processing them, and evolve into an OUTPUT that is in constant evolution.

Physiology is a set of electrical-chemical-physical phenomena that govern the operation of a living organisms. These phenomena must be suited to the conditions in which the body is located. The organisms living in extreme conditions fit their physique in order to survive. The skin and fur of polar animals protects them from cold, as well as large nostrils of the animals that live in hot areas facilitate the cooling of air during breathing. Likewise, individual behavior and social habits of the populations are adapted to conditions in which they are located. The complex system of social hierarchy known in the communities of insect-like ants and bees, in which each individual participates with a very specific role in the survival of its population, is a way to use individual and collective intelligence as a means of adaptation.

This research aims to develop a sustainable architecture method based on a few specific steps: use of parametric design tools to generate single components and larger populations forming architectural surfaces, physical testing of the system through digital fabrication techniques, implementation of the physical system with electronic nervous system, elaboration of an individual and social logic that could act as the artificial intelligence of the system, and test possibilities of adaptation in different environment scenarios, including evaluation of energy balance.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: