What is design if not a solution-driven process? Design, and more specifically architectural design, is an output derived from trial-and-error of various possibilities that are defined by a set of inputs. Everyone would have had a design encounter at least once in their lives. The long road from listing requirements to achieving the final design is one that is riddled with ‘ifs’ and ‘hows’. In the end, are we completely satisfied with the output? What if we had let go of that one constraint that dictated the way other variables shaped up? What if we had chosen the other option over this one? What if we had taken just a little more time to explore other alternatives? Is this the most feasible option for my design? It is normal for questions like these to arise after we decide on something. The ‘what ifs’ need not be a dreaded occurrence if there was a way to test out all our possibilities beforehand.
Generative Design does exactly that. Wikipedia defines Generative Design as “an iterative design process that involves a program that will generate a certain number of outputs that meet certain constraints, and a designer that will fine-tune the feasible region by selecting specific output or changing input values, ranges and distribution” (Generative Design, n.d). Simply put, it is a strategy that teams up a computer program that explores infinite possibilities and a designer who filters from them to arrive at the most suitable choice. It combines the efficiency of AI with the skill of the human mind. Where it is almost impossible for the brain to calculate every permutation and combination in short durations, software can greatly reduce processing time while producing optimal results. Generative design can mean huge savings in human resources, as the designers are spared from having to try out every alternative and can confine their involvement to decision-making.
Generative Design is not a New Concept
It has been trying to make a mark across a range of design disciplines since the 70s but has mostly fallen short as a mainstream revolution. As technology becomes more accessible and affordable, demand for more efficient & innovative systems has allowed generative design to become an accepted approach in construction & architectural design. One of the most famous examples of Generative Design, created by the notoriously futuristic architect Zaha Hadid, is the Heydar Aliyev Center in Azerbaijan. The building has become a landmark in Baku (capital of Azerbaijan) since its opening in 2012. Its architecture shuns angles, replaced by soft curves that arise from the site’s topography and envelop various spaces on the inside. Due to the absence of walls, the complex folds are integral to defining interior functions as well. One look at the structure reveals its perfect geometry to be a winning alliance between computer-aided engineering and habitation sensibilities of the architects. Hadid was renowned for her penchant for using modern technology to create spellbinding shapes; it is of little surprise that she was able to adapt early generative design features of software like Autodesk and Rhino to achieve her ambition.
Autodesk itself has incorporated its generative design tool for the planning of its office spaces in Toronto, launched in 2017. To understand what goes into this form of designing, we need to familiarize ourselves in the broad stages of the process:
- Generation
- Analysis
- Evaluation
- Evolution
- Exploration
- Adoption
The first step is to input pre-determined parameters like occupancy numbers, daylight percentages, proximity of various functions, etc. Data can be limited to few people or can be taken from thousands; the more the information, the more the results are fine-tuned as the program eliminates impractical alternatives. Autodesk gathered information from over 250 of its employees to guarantee that the final floor plan was one that met every worker’s needs. The tool generated over 1000 layout options, each with performance statistics for the denoted data, making it easier to discard low-performing options. Designers, after careful evaluation, refined and further developed them to hone in on the one perfect solution that worked best. A prime example of a powerful machine-human intellectual merger, this form of designing applies to any design challenge, be it planning, form development, or structure-related.
Design in India
Most architects the world over are always looking to embrace novel software and design methods that can reduce workloads and speed up processes. Due to its functional nature, generative design has spanned borders & made its way into the remotest computer screens. Although still in its nascent stages, as key software developers continue to introduce new tools, generative design finds use in many projects, small or large scale likewise. Traditional methods of design can restrict designers in terms of forms they can achieve, number of planning configurations they can explore, etc. With Vaastu being a key component in the residential sector, architecture can often become saturated with cookie-cutter designs. Generative design offers a potential solution to process initial data and come up with unique designs that are project-specific.
The proposed mega-project of Shirdi Sai Baba Temple on the outskirts of Chennai, conceived by Shilpa Architects in collaboration with rat[LAB] studio, has found its origin in generative design and parametric architecture. A modern interpretation of traditional South-Indian temple forms, the geometry evolved from numerology and Vaastu requirements of the client while adhering to spatial and structural constraints. The pre-defined parameters were duly processed by algorithms with respect to contextual factors such as light, wind, terrain, etc. Fine-tuning by the design team led the form to grow from a hendecagon floor plan and blossom into a marvelous 3-dimensional rendition of ancient culture.
Generative Design is a Tool, not The Ultimate Decision-Maker
When using advanced software tools, it is important to remember that these are enablers to a smoother design process. Although generative design can greatly lessen human resources involved in trial-and-error stages, the chosen design is far from complete without adequate evaluation from the architect. Moreover, selecting from countless options is not an easy job; it takes time, patience and a steady mindfulness. Tradeoffs need to be made in deciding whether time would be better spent exploring a handful of select ideas manually or analyzing hundreds of them on computers. After all, developing a few good designs is better than many average ones. The input of data has to be done with precision as slight errors can cause whole analyses to be incorrect and having to redo the tedious processing stage.
Generative design can change the way architects work if adapted properly and only where necessary. Using it as a replacement for architects will not automatically give the best results. For any well-functioning design, human thought & comprehension is a must. While software can help us get organized, only people are capable of deciding what to place where and how to do it. Software gives us the freedom to make informed choices and empowers us to create the best version of a design.
References
Souza, Eduardo (23 Apr 2020). How Will Generative Design Impact Architecture? [Como o Design Generativo deve impactar a arquitetura?] – ArchDaily. [online]. Available at: https://www.archdaily.com/937772/how-will-generative-design-impact-architecture. [Accessed 10 Sep 2021].
en.m.wikipedia.org. (n.d.). Generative design – Wikipedia. [online] Available at: https://en.m.wikipedia.org/wiki/Generative_design [Accessed 10 Sep. 2021].
Designingbuildings.co.uk. (2020). Generative design. [online] Available at: https://www.designingbuildings.co.uk/wiki/Generative_design [Accessed 10 Sep. 2021].
www.youtube.com. (n.d.). Generative Design in Architecture | In the Innovation Zone at AU with David Benjamin. [online] Available at: https://youtu.be/TmAY6qycBsk [Accessed 10 Sep. 2021].
Editor (2021). The Generative Design for Architecture – Things to Know. [online] Go Smart Bricks. Available at: https://gosmartbricks.com/the-generative-design-for-architecture-things-to-know/ [Accessed 10 Sep. 2021].
Baldwin, Eric (25 August 2020). India’s New Parametric Temple to Reinterpret Vernacular Design in Koppur – Archdaily. [online] Available at: https://www.archdaily.com/946233/indias-new-parametric-temple-to-reinterpret-vernacular-design-in-koppur. [Accessed 10 Sep 2021].
archistar.ai. (n.d.). 4 examples of generative design in action. [online] Available at: https://archistar.ai/blog/4-examples-of-generative-design-in-action/. [Accessed 10 Sep 2021].
Davis, D. (n.d.). Generative Design is Doomed to Fail. [online] Daniel Davis. Available at: https://www.danieldavis.com/generative-design-doomed-to-fail/. [Accessed 10 Sep 2021].
