Codify

Parametric and Computational Design in Landscape Architecture

Bradley Cantrell, Adam Mekies, Bradley Cantrell, Adam Mekies

• Publisher: Routledge
• Year: 2018
• ISBN:
  9781317299073
• Edition: 1
• Topic: Architecture
• Subtopic: Urban Planning & Landscaping

Codify: Parametric and Computational Design in Landscape Architecture provides a series of essays that explore what it means to use, modify and create computational tools in a contemporary design environment. Landscape architecture has a long history of innovation in the areas of computation and media, particularly in how the discipline represents, analyses, and constructs complex systems. This curated volume spans academic and professional projects to form a snapshot of digital practices that aim to show how computation is a tool that goes beyond methods of representation and media. The book is organized in four sections; syntax, perception, employ, and prospective. The essays are written by leading academics and professionals and the sections examine the role of computational tools in landscape architecture through case studies, historical accounts, theoretical arguments, and nascent propositions.

Syntax 01

---

Computation in landscape architecture requires that designers extract and abstract rules for landscape performance from an ecological, formal, cultural, and material perspective. Developing the rules, or syntax, of landscape is its own design project, requiring that landscape architects cleverly define how an environment is represented with behaviors and relationships.

01.00 Computation in practice: an inquiry into the business of computational design Jared Friedman and Nicholas Jacobson PAGE 39	01.01 Generative modeling and the making of landscape Chris Reed PAGE 50
01.02 Twenty-first-century learning Pete Evans PAGE 64	01.03 The parametric park David Fletcher PAGE 71
01.04 Big data for small places: systematizing the functional use of complex and sizable data sets in daily practice Elizabeth Christoforetti, Will Cohen, Yonatan Cohen, Stephen Rife, and Jia Zhang PAGE 77	01.05 Turing landscapes: computational and algorithmic design approaches and futures in landscape architecture Stephen M. Ervin PAGE 89

01.00
Computation in practice

An inquiry into the business of computational design

Contributors:

Jared Friedman
Computational designer and licensed architect

Nicholas Jacobson
Designer at Davis Partnership Architects

FIGURE 1.0.1

Image: Paralabs

The terms “computational design” and “parametric design” can be defined in many ways. They may bring to mind forms driven by generative algorithms or the ability to design with various types of data sets or simply the use of Building Information Modeling (BIM) software. We can argue that what the fields of computational design and parametric design hold in terms of potential and capability, they lack in clarity and specificity. How do we begin to define these terms to a potential client? And how do we choose which computational juice is worth the squeeze? These are just a couple of the considerations when integrating computational and parametric design strategies into practice.

FIGURE 1.0.2 Landscape radiation and environmental analysis. With the property devoid of vegetation, creating habitable outdoor spaces would be a challenge. Thermal and wind analyses informed critical design decisions – each aimed at the goal of extending the hours and seasons during which the landscape can be used

Image: Design Workshop Inc and Paralabs

FIGURE 1.0.3 Process of dress fabrication

Image: Paralabs

A designer that is working parametrically may be thought of as someone who is flipping the traditional process, an editor of constraints first, and an empirical designer once the constraints are designed. What this means for practice is often far more interesting and innovative than the complex forms that initiate such projects. Inherent in the computational design process is the ability to quickly explore multiple design iterations throughout a project. This process is sometimes referred to as “optioneering,” in which consultants and collaborators are brought on early in a design in order to help define the project constraints. As design processes and schedules adapt to emergent modeling and analysis workflows, we find opportunities for new models of practice that simultaneously react to, and influence advances in the field of computational design.

The following interviews were conducted by Paralabs, a small computational design practice, founded in February 2016. Our design lab is built on the notion that quality design emerges from a rigorous approach in which designers are conscious of, and able to work with, more of the parameters at play. We utilize computational thinking and techniques in order to investigate design problems, and ultimately arrive at creative solutions that are informed by the context.

Within the early development of our lab we have found ourselves constantly adjusting our business model to allow for flexibility in exploring the applications of transdisciplinary computational design, where we work to create new intersections with other professions. We have collaborated on projects with fashion designers, medical professionals, and landscape architects.

It is not initially obvious nor easy to explain how a computational way of thinking can be applied to new and unexpected collaborations. Hoping to shed some light on the subject, we have chosen to investigate how other successful computationally focused practices situate themselves, both from a design perspective and a business perspective. All of the questions asked are questions that we have asked ourselves since forming a business. Leaders from four unique design and engineering–based practices have been selected to gather a variety of perspectives and insights on how they integrate computational thinking into practice. We hope that these conversations can also provide a glimpse into the struggles and opportunities for an emerging computational design practice.

The interviewees

Bill Allen (BA)

Computational Designer
Partner and Director of Building Information
Management Services at EvolveLAB

Bill has over a decade of experience managing technology for buildings in the AEC industry including managing large, complex BIM workflow mechanisms as a BIM manager for cutting-edge firms. He is an expert in parametric systems and interoperability between different software. EvolveLAB is a full-service BIM consulting firm. As building technology professionals, they influence and mentor engineers, contractors, owners, and architects. Their mission is to make project teams more successful through innovation, progressive workflows, and technology. They provide the tools, the resources, and the knowledge to make project teams more efficient and sustainable.

Hanif Kara (HK)

Structural Engineer
Founding Partner of AKT
Professor of Architectural Technology
Harvard University

AKT is a London-based progressive designled structural and civil engineering practice. P.ART is their nonprofit in-house parametric research team working to develop bespoke software programs to interrogate and analyze projects at any stage of the project.

Benjamin Koren (BK)

Managing Director of ONE TO ONE

ONE TO ONE is a computational geometry and digital fabrication consultancy on art and architecture projects, offering services in bespoke geometric computation, precision 3D CAD construction, integrative CAM fabrication and innovative R&D at all scales.

Andrew Witt (AW)

Co-Founder of Certain Measures
Lecturer at Harvard University

Certain Measures is an office for design science using geometry, mathematics, data, and new technologies to create new spaces and ways of making. Tapping open-source software and knowledge, embracing a flexible collaboration structure, and building new software and robotic tools for these novel projects, Certain Measures draws on diverse backgrounds in architecture, robotics, computer science, and mathematics, to simplify beautiful complexity and to delight, surprise, and enrich human experience.

Interviews

Paralabs (PL): Computational design is a somewhat new process that can be foreign to clients. How is marketing this process different than a traditional design process? How do you explain this process to potential clients?

AW: In and of itself, computational design is too vague and, to most clients, abstract to be a useful category. First, it becomes this kind of Rorschach test for what potential clients think computational design is, which can be very limiting. Second, it becomes too easy for clients to think computational designers all have the same skills and are simply interchangeable. What matters is the specific domain expertise or your specific communicable perspective, which is extended by computational tools.

We don’t think of Certain Measures as a computational design practice but as an office for design science or as an incubator for design technologies. In this sense, we focus on particular domains – such as machine learning, urban optioneering, and of course design geometry – around which we develop our own specific products and technologies. We solve problems which are not computationally defined but which admit computational solutions.

HK: I’m not convinced it’s different. It’s current and relevant. It’s a way of adding value to everything you are thinking about. It could be losing weight, better design, or making something cheaper and quicker. Manufacturing helps to collapse a process down the line. Sadly, some people call it BIM – it is progress, but it’s limited. When we didn’t have projects, we showed what the potential was, and there is an element of marketing through showing prototypical ideas. Later on we did buildings that proved these concepts.

FIGURE 1.0.4 Close up view of sound panels

Image: Ben Koren, ONE TO ONE

BK: It depends on the client. To architects and artists, we stress the quality and precision of our work and the potential it offers in enhancing their vision. To workshops and contractors we stress efficiency, speed and cost reduction in using automation and optimization techniques. In the best projects, it is a balance of the two, when working on both ends. However, we tend to avoid the term “computational design” so as not to compete with the creative work provided by the designers. We understand our practice to be a consultancy firm in computational geometry, not a design firm.

PL: Digital tools allow designers to collaborate in new and different ways. How is your business management model and interactions with collaborators different than a traditional design practice?

AW: At Gehry Technologies, I developed a prototype for a product called GTeam (now Trimble Connect), which was all about new modes of digital collaboration. While it was a fascinating process, what I found was that it is a heavy, slow, and heroic effort to try to change business processes, particularly when you are only a consultant. It requires extensive training and a real change in mind-set by many stakeholders. Quantifying the benefits those changes would bring is also a challenge. I’m glad someone is doing it, but I think there are more effective ways to provide value. In short, it is the opposite of the kind of problem around which you can build an efficient business.

Now we try to focus on methods and processes that are “productizable,” that is, they can be packaged and consumed with as little friction as possible, and produce impactful results without boiling the ocean. It also makes it much easier for these tools to be sold or spun off later, because there is not a heavy service component.

HK: The business model has not changed. The way we collaborate is through being able to do more work early on. With architecture, we can very quickly send data and optimize before we even present to a client. In the old days an architect would do a presentation and the engineer would respond, then the client would say yes or no. These days we are able to do most of that computationally. So in terms of what is expected of us, early, what we find is that clients are expecting a lot more refinement because they understand the design better. I would separate this from “optioneering” because that is just drawing prototypical ideas for all kinds of types; but when you are engineering properly, and optimizing with an architect, you can actually optimize at many levels because it has to do with light, environment, structure, and architecture – not just structural engineering and architecture.

The business model isn’t different and the fees don’t go up hugely just because nobody wants to pay any more; they just want more for their money. The difference is this: after building a reputation in it, your initial number for your fee is always higher. So when people call you in the first instance, they know they are going to pay 10 percent or 15 percent more than calling the guy down the street because they know that the quality of service they are going to get is much higher, and the return that they get in their value – whether that is lighter, quicker, etc. – is worthwhile.

The investments we made in computation were not intended for immediate return. It took time while the culture was being built; and when you see the response that we produce from having built that investment, t...