Generative Design: algorithm-driven engineering and architecture
“The harmony of the world is made manifest in form and number, and the heart and soul and all the poetry of natural philosophy are embodied in the concept of mathematical beauty.” So wrote the British mathematician and biologist, D'Arcy W. Thompson, in 1917 in his On Growth and Form. A hundred years later, generative design is putting these words into practice. But what is generative design? It is a branch of design that makes use of algorithms and artificial intelligence to process complex mathematical and geometric parameters to obtain a digital model of the solution required. “To explain what I am talking about, I often use the metaphor of the flock,” says Giorgio Buratti, Professor of Digital Space Modelling at the Politecnico di Milano. “The form of a flock of birds in flight results from the distance and position of each bird compared to its neighbour. These distances and these positions,” continues Mr Buratti, “are simply the parameters that a designer enters into a software in order to obtain the form”.
From aircraft wings to biomedical prosthesis, through to the texture of a particular fabric or the realization of a particular architectural solution, the applications of generative design are practically endless. And yet, the promoters of this minor revolution are a handful of engineers, designers and architects who are unafraid of abandoning traditional paper sketches in favour of digital design. A community that revolves around software packages such as Grasshoper (an extension of Rhino), Autodesk Within and Fusion 360, to mention a few. Although these names don’t mean much to most of us, they have the ability to turn a virtual project into reality, through the possibility of interfacing directly with a 3D printer or a CNC machine. “Generative design is closely linked with rapid design systems. In 2002, at the time of my first show, there were only two 3D printing systems in Italy. Now things have changed, and we are facing a brand new multiplier effect”, says Marco Maggioni, Professor at IED, Milan.
Some high schools have already introduced courses is computational design. The aim is to develop a logical approach to problem solving.
Mastering this design system, however, requires a mental shift: we must no longer think in terms of form; instead, we need to select the correct parameters to be used, from which form then emerges. This process requires a good dose of mathematics and geometry. “It’s a pretty steep learning curve at the beginning,” says Mr Buratti. “Although the calculation is performed by the computer, we have to enter the initial data, and if this isn’t suitable we won’t get very far. It’s a bit like going on a date. When you’re meeting someone, you give them directions: I’ll meet you at such a time, in such a place, which you can get to by taking such a bus. Taken together, this information is the algorithm you will use to kick off the project you have in mind”. What matters, like on a date, is to be original. “This kind of software is often used badly,” says Mr Maggioni. “Without a hypothesis and a precise objective, you risk running up against the sterile production of trite designs.”
Consequently, even more important than scientific expertise are values such as curiosity, a desire to experiment, and a predisposition for analysis and research. All this goes to supplement a training path that, for those studying design at university, already includes a knowledge of modelling rules, design techniques, materials and how they are processed. In this regard, Italy is finally taking steps forward: “Some high schools have already introduced courses is computational design. This is a great result, if you think – as the saying goes – of Italians as a people of saints, poets and sailors. The aim is to develop a logical approach to problem solving,” explains Mr Buratti.
General Electric, for example, has used generative design to create an aircraft engine. The former one was made up of nineteen pieces. The new one, instead, was printed in 3D, and consists of a single piece that weighs a third of its predecessor. Similarly, Airbus has designed a new panel to separate hostesses and stewards from the passengers. The result? The new panel weighs 55% less that the original one. And airline customers are grateful for the savings. Basically, where there is complexity, generative design offers cutting-edge solutions that would not be possible with pen and paper.
So, will algorithms steal designers’ jobs? There are those who are convinced they will, and those who are sure this will never happen. “Generative design is a kind of invisible revolution,” explains Stan Przybylinski, Vice President of Research at CIMdata, an American consulting firm. “It improves project quality, efficiency and performance at all levels as it optimises the choice of materials and generates infinite variations.” But the concept and form will always be the prerogative of the designer, who will be able to choose between these infinite variations: “The inside of a moka pot, for example, will be developed by a computer, and the outside by a designer,” he explains. Nothing more than a helping hand, therefore. If only all revolutions were like that.