Home / Blog / Design / 2 Luglio 2021Generative Design and the New Frontier of Design
One Type of Generative Art: Computational Design
The new frontier of design is called generative design, which arose from the idea of using artificial intelligence—the branch of computer science that studies the development of hardware and software systems capable of independently pursuing tasks and making decisions usually entrusted to Man—in support of the designer.
Generative design can be theoretically applied to any type of ideation process: it is more productive, however, in terms of technical-economic feasibility in the case of architectural design. Architecture professionals are increasingly applying themselves to scientific research related to digital design. Computational design therefore means a new approach that sees the replacement of the representation of the project, be it an object or a building, with its simulation: the design of individual objects is moving towards the use of integrated and complex systems. If the architect’s work was previously relegated to intuition and experience, in the 21st century it can count on using programming languages.
This new working method allows the design and creation processes to be speeded up and optimized. Regarding this last point, the doubt may arise that the designer’s imagination is overwhelmed by the efficiency of an algorithm. In reality, the designer is tasked with setting objectives and constraints relating to the project, which will then be analyzed by the computer using algorithms. At that point, thousands of options and provisions will be generated by the computer itself, without this entailing the passivity of the designer, who, in turn, will be evaluating alternatives or modifying them.
Generative Design between Designer and AI
There is, therefore, a strong synergy between the designer and artificial intelligence: the designer’s creativity is not suppressed. On the contrary, it is precisely encouraged thanks to the conception of a large number of possibilities and options to be experimented and explored.
In Generative design, as in generative art generally, the goal is the process and not just the final result. The designer pays to the design process, or rather, to process monitoring. During this phase of transformation and innovation, the designer will not lose his central role but will redefine it in relation to new working methods. The model will no longer be the final goal, but will become laboratory analysis: it will require architects to think about which is the best realization in relation to the conditions and the context, no longer working on this or that specific building project.
The Origins of CAD
Generative design is the result of the development of CAD (Computer-Aided Design) technologies—using computer graphics to support technical drawing—and is therefore a new approach to design that involves computer assistance in the creation of the models on which the designer wants to work. Design is then accompanied by Computer-Aided Manufacturing (CAM), which uses geometric design data to control automated machinery.
CAD originated from three different sources which highlight its basic operations: the first concerns the result of attempts to automate the process of creating and editing the project with the advantage, compared to traditional drawing methods, of being able to correct computer modeling much more quickly by changing its parameters. The second source relates to the verification of designs by simulation, and the third source of CAD development reflects the result of efforts to facilitate the transition from design to production.
Today, the evolution of CAD techniques is moving towards a full integration between the physical and the virtual model: the improvement of virtual models is a fundamental step in the evolution of the industrial landscape which, thanks to the use of simulation, has been able to learn to predict the physical response of the model under certain conditions of production or use.
Another advantage of generative design is the ability to consolidate parts: given that generative design can handle a level of complexity impossible for human engineers to conceive, and that additive manufacturing—the industrial process used to manufacture objects from computerized models, adding one layer on top of another—allows for fabrication of complex geometries produced by generative algorithms, it is possible to create individual parts that can replace assembled parts and this allows supply chains and maintenance to be simplified, as well as reducing production costs. In fact, thanks to the design, simulation and manufacturability solutions, generative design has a greater impact than traditional design: it affects the entire production process.
Parametric Modeling and Algorithms
The designer who works with computational design has the task of entering data, or parameters, into a software. For this reason, computational modeling is also called parametric modeling: a type of three-dimensional modeling that is based on relating various components of the model to each other with what are, in fact, called parameters, which are the numerical values aimed at defining a characteristic of the object represented. Parametric, in this specific case, is synonymous with algorithmic: parametric modeling software exploits mathematical algorithms to describe the processes which, in turn, are translated into three-dimensional models.
The use of parametric modeling for design makes it possible to overcome geometric limits and obtain unique solutions: once the project of the object has been defined, the parametric definition is written, specifying the dimensional, quantitative, material, and formal codes on which to act, generating completely new shapes and ideas, producing geometries that would be almost impossible to obtain with common CAD tools.
Collaboration between Algorithms and Engineers
The Generative design process actually aspires to guarantee a more efficient collaboration between computer and user, using algorithms based on Machine Learning principles, such as genetic algorithms.
It is, in fact, an iterative design process, where the designer defines the variables and constraints that he will then formalize in the algorithm, which will subsequently act by generating a certain amount of results. This mechanism is repeated until a product that satisfies the designer is reached: artificial intelligence cannot, then, replace the critical and creative capacity of the human being, it is rather a tool.
Morning Future writes:
“From aircraft wings to biomedical prostheses to the texture of a particular fabric or the creation of a particular architectural solution, the fields of application of generative design are almost infinite.”
The instigators of this revolution are that minority of engineers, designers and architects who have had the courage to abandon classic design and rely on digital design which, to be mastered, needs a certain level of mental agility. Now the work must be turned towards the selection of the correct parameters to be used to bring out the desired shape. In addition to a good dose of mathematical and geometric knowledge, and—more generally—scientific skills, the most important characteristics to approach this new practice are curiosity, the will to experiment and the predisposition to research.
A new landscape opening up for design allows us to reflect on the changing role of human beings and computers in their various fields of collaboration and contact: throughout history, man has shaped the world with design, intended precisely as a technique and act on the world. But now that AI is becoming superior in taking on certain roles within the design process, humans will have new tasks.
In other words, in the future, designers will work less on design. Instead, they will aim at supervising, guiding and setting parameters for computational designs.
The question then arises: how will this new panorama modify the perception of the value that is given to the designer’s “profession”? nd how, once the machines have been given greater design autonomy, will the shape and design of the objects that make up the real world be transformed?
cover image credit: The Met (Object: 208895 / Accession: 1997.518)