Measurement, Uncertainty, and the Dataspace

Authored by Nathan Miller

This article was originally published in Oz Journal v33. The version below contains some additional graphics and modifications to the published text.

Instruments

We shape our tools and they shape us. – Marshall McLuhan

TPG_DesignInfo.jpg

New digital technologies are dramatically transforming our environments and behaviors. Today’s vast communication networks, comprised of over 140 quadrillion transistors, filter volumes of information at rates unimaginable just two decades ago. Individuals occupy the space of the city armed with smart phones more powerful than the computers that enabled the first moon landing. Virtual and physical lives have collided on social networks.

In this context, Marshall McLuhan’s observation about the feedback loop between tools and users is of special importance when considering the role of the architect and technology.

When we study the history of the architect’s technology, we can immediately observe significant gains in accuracy, efficiency, and portability. The move from animal skins to industrialized paper revolutionized duplication of documentation. The adoption of CAD and BIM technologies greatly improved accuracy and coordination.

More so, the digitization of our tools is of special significance that extends beyond the purposes of efficiency. The full force of this transition to the digital has been a slow explosion over the past 3 decades which ultimately gives architects and designers direct control over a new medium: information.

Information, itself, presupposes no specific type of representation. This is an important consideration. 2D drawings and 3D models have represented information in a graphic form. However, other representations such as logic trees, databases, mathematical formula, and algorithms offer a fundamentally different means of designing and producing architecture.

Information requires new set of tools for precise control and manipulation. Algorithms and computer languages, which have existed on the back end of digital user interfaces, are coming to the front of the process as designers begin to craft their digital processes by creating scripts, apps, and plug-ins to control information directly.

Measurement

NBBJ_HANGZHOU_PARAMETRIC.jpg

Whenever I draw a circle, I immediately want to step out of it - Buckminster Fuller

Measurement has become more than a predictable means to a quantified end. With our design and analysis tools, we can computationally measure and quantify just about any aspect of the design in terms of geometry, material, cost, energy consumption, and the list goes on. We can measure old things in new ways and vice versa. Over the course of a design process, we may generate an unprecedented volume of information.

Parametric tools, scripts, and algorithms enable the designer to generate, navigate, manipulate, and evaluate torrents of information in novel and unexpected ways. For example, a shortest-path algorithm can enable the architect to design a new kind of building circulation system or understand complex urban networks. Mathematical formula can be used directly to generate and rationalize expressive geometry into cost effective shapes and forms. Client and user criteria can be encoded directly into a parametric model to control the creation of spatial volumes.

NMILLER_URBAN%20GA.jpg

With parametric software, it is possible to encode comprehensive systems of information and representation. The Hangzhou Tennis Center is an example of a complete parametric system which defined systems for geometry, structure sizing, cladding, analysis, coordination, and documentation in a single graphic algorithm with custom scripts… and it is only 10 megabytes in size.

The possibilities for innovation are quite endless if the design of tools for processing information occurs concurrently with the design of architecture.

Uncertainty

On two occasions I have been asked,—"Pray, Mr. Babbage, if you put into the machine wrong figures, will the right answers come out?" … I am not able rightly to apprehend the kind of confusion of ideas that could provoke such a question. - Charles Babbage

While we can continuously generate and measure information, paradoxically, there are greater levels of uncertainty present in the design processes. The Garbage in-Garbage out maxim holds that the value of our design inputs have a direct effect on the value on our outputs. The clarity of this concept, while still a fundamental rule of thumb, is ultimately reductive in describing the complexity of technology and the world in which we use it. We design with tools that perform calculations far more complex than the average user can understand. Simultaneously, the tools are not yet precise enough to describe, with absolute accuracy, the simplest of natural phenomena.

Garbage in - Gospel out is the emerging phrase which might more accurately describe the technological moment the design world occupies.

Given these scenarios of uncertainty, tools for evaluation and selection are crucial. One computational tool that can be employed to navigate an uncertain information landscape is the genetic algorithm (GA). A GA enables the designer to implement an automated search and selection process for identifying outputs that are of value given a set of inputs. Using a GA, hundreds, if not thousands of design options and alternatives can be explored enabling the architect to discover possibilities in a vast information landscape.

Dataspaces

Only wimps use tape backup: upload your important stuff on FTP, and let the rest of the world mirror it. - Linus Torvalds

Collaborative mediums, such as those found in social networking communities, offer a very different model for collaboration based on a kind of simultaneity not found in design tools. Twitter’s system of commenting (tweets) and sharing other comments (retweets) results in highly complex, relational network of information exchange and collaboration. The information network takes on a life of its own.

Many things prevent the open and simultaneous sharing of information including liability concerns, technological preferences, and delivery methods. A fractured information landscape is the natural result of a complex and diverse industry. Under these circumstances, is an open, collaborative system possible?

Freedom to move information across a complex network of users is often hindered by closed software formats and data translation issues. As these issues are overcome (sometimes with jury-rigged or ad hoc solutions) a ‘dataspace’ emerges which is capable of managing multiple information sources.

Other computation processes and tools offer potential for stitching together tools and teams. Inter-process communication is a means by which different software technologies can begin to communicate and share information directly. Open relational databases are another possibility for stitching together information from multiple data sources into organized schemas.

The process for designing the Hangzhou Sports Park in China required as much attention to the development of the information network as it did the development of the complex parametric geometry systems. The international project had multiple, specialized software tools being used between the design teams and engineers.

It was therefore necessary to construct a road map for how the different teams could work together with their respective technologies. The communication solutions on Hangzhou ranged from using different file export procedures to designing custom scripts and algorithms for automating information transfer into various design and analysis tools.
When taking a macro-view of the process, design solutions are represented as both the development of the architecture as well as the development of collaborative communication systems for transmitting design intent.

Ubiquity

In the course of evolution every technology is put to the question of what happens when it becomes ubiquitous? What happens when everyone has one? - Kevin Kelly

While the AEC industry is no stranger to digital tools in the design and production process, the full effects of accelerating information-based processes are still yet to be realized in the form of a comprehensive ‘paradigm shift’.

Architects and designers have become quite enamored with the ability to generate new and exciting forms and processes with scripts and parametric tools. Indeed, it is possible to design just about any shape or form imaginable. The line between the software developer and the user is blurring: the designer codes their tool as they conceptualize and develop their design.

What is typically ignored, however, are the broader effects on an industry-wide scale. What happens when a script becomes a deliverable for use by the construction worker? What happens when a client is given control over the design parameters so they can push and pull the project to their liking? Who is the designer if all information is built on a collaborative intelligence?

The full impact of the computational design ‘revolution’ is yet to be realized… but the transformations to the industry we are seeing today is nothing short of revolutionary…

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License