Computational Design
Brendan Harmon
Welcome.
I am Brendan Harmon,
an assistant professor
of landscape architecture
at Louisiana State University.
This lecture is an introduction
to computational design
and how it is tranforming the practice
of the landscape architecture.
Algorithms
“algorithm, n”. In: Oxford English Dictionary Online. Oxford University Press, 2022. url: https:
//doi.org/10.1093/OED/1019775631.
Algorithms are “precisely defined sets of mathematical or logical operations for the performance of a particular task”. When used creatively in art and design algorithms can evoke a peculiar aesthetics of infinity.
Generative Art
Algorithms that model the behavior of natural systems, such as fractals, cellular automata, branching and venation, differential growth, hyphae, slime molds, and flocking, can produce infinite, complex patterns that evolve over time. In generative art, artists use such algorithms as performances that hide, reveal, and transform meaning. As the algorithm runs, step by step, the performance unfolds. Although the code of the algorithm may remain hidden, the artist’s intention is transformed by the algorithm; the artwork’s meaning accumulates through performance. Even when the audience only sees a final resulting image or form – rather than an ongoing animation or performance – the algorithm leaves its mark, traces of its process, of its logic. Algorithms can express infinity through endless performance and the emergence of complexity from simple rules. While works of generative art and computational design are diverse, spanning a wide range of disciplines, media, and themes, there is a common current – the aesthetic of the algorithm. This algorithmic aesthetics of infinity is abstract, non-representational, and performative.
Jared Tarbell's Infinite Regeneration particles drift endlessly through a field of procedural noise, evoking the emergence of complexity from simple systems.
In Quayola’s Pleasant Places, film footage of a landscape is gradually transformed by algorithms to reveal the sublime in nature. As image analysis, image processing, and audio signal processing algorithms progressively distort the original footage and sound, the landscape becomes more and more abstract with windswept leaves, grasses, and flowers gradually transforming into swirls of color. In its eventual abstraction this video alludes to the impressionist tradition of landscape painting, but as a performance it explores the dynamic nature of sensation, memory, and the environment.
In Quayola’s Pleasant Places, film footage of a landscape is gradually transformed by algorithms to reveal the sublime in nature. As image analysis, image processing, and audio signal processing algorithms progressively distort the original footage and sound, the landscape becomes more and more abstract with windswept leaves, grasses, and flowers gradually transforming into swirls of color. In its eventual abstraction this video alludes to the impressionist tradition of landscape painting, but as a performance it explores the dynamic nature of sensation, memory, and the environment.
Quayola’s Remains are a series of point cloud renderings of woodland scenes, captured with
laser scanning and rendered as points. An exploration of light as a medium, these point cloud renderings are both indexical representations of reality and algorithmic abstractions. The imperfections of the technology give rise to a painterly, pointillist aesthetic that evokes the material and immaterial.
Computational Design
in Architecture
Form finding
Simulation & analysis
Digital fabrication
Autonomous construction
In computational design, this aesthetics of the algorithm is given form in landscape and the built environment. In computational design, designers, rather than drawing or modeling a singular form, write code or compose visual programs that generate form. The designer scripts a creative process, a generative system. By designing algorithms rather than discrete forms, designers can rapidly explore design ideas, generate variations on a theme, run simulations, and evaluate performance. In architecture, computational design is used for form finding, simulation and analysis, digital fabrication, and autonomous construction.
dECOi Architects, 2011
Digital fabrication, by giving architects more control over the construction of their designs, has spurred structural, material, and aesthetic explorations in architecture. Architects use technologies such as laser and plasma cutting, CNC routing and milling, 3D printing, and robotics to build prefabricated parts for complex assemblies. Projects such as dECOi Architects’ interior for One Main used CNC milling to carve unique wooden components for organic, parametrically modeled forms.
The interior of One Main was assembled from non-standard, CNC milled panels of sustainably harvested spruce plywood to create a novel, sinuous aesthetic in which architectural elements and furniture blend together as a cohesive whole.
Michael Hansmeyer & Benjamin Dillenburger, 2017
With digital fabrication, the aesthetics of the algorithm can be expressed through built form.
Hansmeyer and Dillenburger's Digital Grotesque, for example, evokes the infinite.
This form was generated by algorithm using volumetric modeling and was 3D printed in sandstone.
Michael Hansmeyer & Benjamin Dillenburger, 2024
Architects have been experimenting with robotic processes for autonomously constructing buildings such as laying bricks, stone cutting, assembling timber frames, constructing complex formwork, winding tensile structures, and 3D printing metal, concrete, and soil structures. For example, Tor Avla by Hansmeyer and at ETH Zürich is being constructed from pre-fabricated columns 3D printed concrete
When robots, programmed by architects and engineers, construct buildings components, the design team has direct control over fabrication. With robotic fabrication and construction, architecture becomes performative; architects are no longer just designing form, but rather the process by which form takes its shape.
Gramazio Kohler Research, 2011
While digital fabrication has extended architectural agency from the design office to the machine shop, robotics promises to extend architectural agency to the construction site. Pioneering work by Gramazio Kohler Research at ETH Zürich explored the creative, aesthetic, and material implications of on-site robotic construction.
In the Endless Wall, an articulated robotic arm on a mobile base tirelessly performed the Sisyphean task of continually reassembling a brick wall. As a performance, the Endless Wall expressed an algorithmic aesthetics of infinity through its continual cycle of construction and deconstruction.
MAEID, 2018
The interdisciplinary practice MAEID explores the entanglement of technology and nature through creative experiments in robotic biofabrication such as Pahoehoe Beauty, Magic Queen, and Sylva.
In MAEID's Pahoehoe Beauty,
a landscape is 3D printed with mud.
With its strange techno-organic aesthetic
- defined by the biologically-inspired
algorithms for dendritic growth and the
physics of mud -
this work of speculative design
problematizes binary conceptions
of technology and nature
by entangling them in uncanny ways.
MAEID, 2021
In MAEID's Sylva,
robotic gardeners tend
an artificial ecosystem.
The robots, using machine vision to interpret their environment, are programmed to seed, water, or trim the plants. While the plants rely on the robots for care, the robots – acting in response to their environment – only move in response to the plants. This performative installation asks how technology and nature can coexist in symbiosis, exploring themes of empathy, coexistence, non-human subjectivity, and distributed agency.
Computational Design
in Landscape Architecture
Form-finding & simulation
Autonomous planting
Autonomous earthworks
Remote sensing
Landscape architects use algorithms to generate paving patterns, planting patterns, and landforms. With remote sensing and robotics, they can autonomously map, construct, and plant these complex, algorithmically generated designs.
Land Collective, 2017
For the Cummins Distribution Headquarters...
Land Collective used an algorithmic approach to rapidly generate alternative paving patterns from fading gradients to the final design of alternating stripes of light and dark. This enabled the design team to quickly iterate through design concepts, testing many different ideas. The aesthetic is computational in its legible complexity, its sense of movement, and its nod to pixels and barcodes.
Il Nature, 2015
At Il Nature's Les Jardins d’Etretat...
Procedural modeling was used to design the extensive topiary,
which are meant evoke the energy and dynamism of the ocean waves
below this cliffside garden.
Snøhetta, 2016
Computational design can also be used to build high performance landscapes. Snøhetta designed the MAX IV Laboratory Landscape for functional performance, algorithmically modeling and autonomously constructing landforms that would dampen ground vibrations around the particle accelerator and collect stormwater.
GPS-controlled bulldozers were used to precisely construct these complex landforms.
Benedikt Groß, 2013
For Avena+ Test Bed, Benedikt Groß designed algorithmic patterns of crops, herbs, and flowers for wildlife friendly farming that were planted using a tractor with GPS-guided machine control.
Using a Voronoi diagram, an eleven and half hectare field was partitioned into cells planted with oats and bordered with herbs and wildflowers to attract pollinators and predators for pests.
Jud, Hurkxkens, & Hutter, 2020
Reseachers and designers at the Swiss Federal Institute of Technology in Zürich...
...used a legged robotic excavator to autonomously construct an embankment.
The embankment was built using on-site material, balancing cut-and-fill.
Here, construction becomes on ongoing performance directed by the design team and enacted by machine.
Christophe Girot, 2019
“designers adopting cloudism will step into an overwhelmingly convincing simulacrum of physical reality, space and time; this will enhance their understanding of site, and yield a stronger awareness of ambient aspects and cues”
https://doi.org/10.1515/9783035622164-013
While remote sensing technologies like lidar have long been used by landscape architects for mapping and modeling terrain, lidar point clouds are emerging as a new media for the discipline, with great analytical, phenomenological, and aesthetic promise. Christophe Girot wrote that "designers adopting cloudism will step into an overwhelmingly convincing simulacrum of physical reality, space and time; this will enhance their understanding of site, and yield a stronger awareness of ambient aspects and cues."
Philipp Urech, 2019
Sections cut lidar point clouds can reveal the ecological structure of a landscape.
In this example from Singapore, the ground has been re-regraded and the existing plants fit to the transformed terrain.
Philipp Urech, 2016
In this example from Dublin...
an airborne lidar dataset of the city is transformed to envision a green retrofit.
