The communication that constitutes a project is, however, still typically carried out between human-to-human actors. As computing becomes more pervasive and more embedded into our physical environments, buildings will begin to communicate information of their own. In this way, the AEC industry will begin to echo the software development industry: in that information will be mined from every aspect of our projects. But whereas As the built environment goes through a process of data-fication, what does a world of bytes mean for the architect?the information mined from software products is user clicks, traffic, algorithmic efficiency etc., buildings products will analyze environmental data, user-space interactions and higher level qualitative metrics.

We refer to this process as the Quantification of Space, which heralds a new era in spatial analytics. An era marked by evidenced-based and data-driven decision making, user-centric concerns and the ability to inform design iterations through new modes of human-machine feedback. As spatial analytics becomes more prevalent, sophisticated, and in turn multidimensional, architects will need to critically address the questions and challenges that these new horizons present. As the built environment goes through a process of data-fication, what does a world of bytes mean for the architect? When anything is measurable, what are the meaningful metrics for the architect to measure? And how does this affect the formation of an architect and architecture’s role within society? We endorse the quantification of space as a method that allows for a new type of awareness of the built environment at large. 

A new form of quantification

To describe qualities of space through quantitative units is to go beyond the subjective assumptions and firsthand or lived experience of it. To break down high-level qualitative attributes into discrete units of raw metrics, is to re-read a textural landscape, and provide unexpected insights into the environments we inhabit. What architectural features and spatial organizations promote social interaction in a bar, for example? Or how might one describe the feeling of comfort in a hotel through the variance/standard deviation of room area? Mining our spaces for seemingly unrelated data sets, enables us (as designers) to uncover hidden relationships—whether causal relationships or not—between space and lived experience.

We are against measuring for measuring’s sake or counting spatial features for a cold-hearted optimization algorithm.The notion of describing space through quantities, however, is not new. It can be traced back to the Space Syntax theories and techniques of the early 1980s that aimed to analyze the complex spatial configurations of the built environment. More recently, BIM has enabled a further number-driven perception of space in which every aspect of the physical environment has a byte representation. Thanks to the rapidly expanding world of the Internet of Things, the possibility of measuring spatial attributes and environmental phenomena has expanded. It is within this lineage of analytics and advanced computing that we locate and define an alternative to this notion of the quantification of space. We are against measuring for measuring’s sake or counting spatial features for a cold-hearted optimization algorithm. Instead, we shift our focus from space itself to spatial experience, with the intention of quantifying the possible correlations between space, user behavior, and fine-grained phenomena.

While Space Syntax sought to use an inductive methodology to analyze the “configurationality” of space and identify the “un-discursive relational schemes that structure” the built environment, we seek to achieve a non-linear interplay between inductive and deductive reasoning to quantify high level, design-specific qualitative spatial attributes and experiences*. From a general hypothesis or metric about a spatial phenomenon, we are able to derive a set of lower level sub-metrics, which serve as observation to challenge the initial statement or infer an entirely new type of generalization. 

Beyond data: space analytics as a learning process

New (expanded) landscapes require new methods of legibility. This formed the basis for “Measure,” a graduate seminar co-taught at Columbia University’s Graduate School of Architecture, Planning and Preservation (GSAPP). The students were charged with developing a hypothesis about the spatial aspects of an existing site in relation to larger non-spatial phenomena; to invent alternative forms of measurement in order to generate new metrics to evaluate and judge the spaces in which we operate. Students were asked to in an attempt to construct an equation about spatial experiences, designers must not only identify their biases, but also explicitly assume a stance towards design and the use of space.broadly rethink measurement apparatuses: the filter or methods in which information is gathered and knowledge is generated. Using computational design tools and custom embedded computing devices, students attempted to extract, manipulate, and visualize information about the experience and spatial qualities of an environment in order to develop new units of measurement, and reveal more obscure relationships between data-sets. The goal of each project was to form critical arguments about the use or systems of space.

The seminar attempted to go beyond data and statistical analysis in order to unearth the impact of spatial, social, and behavioral aspects of certain qualitative metrics. One such example from our graduate course was a project that hypothesized the relationship between anxiety and one’s position within an architecture studio: that the level of connectivity or “betweenness” of an individual's seat and position has a direct impact on her or his degree of anxiety. The project asked, “could anxiety levels be understood from the frequency of a person’s movement at their desk, the amount of people walking behind that same position, and the arrangement of neighboring desks?” The point of the class, and of this project, was not to scientifically validate or falsify the legitimacy of the claim—that anxiety has a relationship to movement—but to force designers to study the interdependent relationships between a space’s configurationality and the wider sphere of user behavior and experience.

Indeed, in an attempt to construct an equation about spatial experiences, designers must not only identify their biases, but also explicitly assume a stance towards design and the use of space. Thus, the act of quantification becomes also a form of learning.  

The course assumes that space is a living organism that needs to be monitoredThe very absence of this subject in contemporary architectural education can be attributed to the still prevalent belief that the architect is the central author in a project—a belief that states his or her role is complete, or stops, once said building is “complete.” “Measure” attempts to disrupt the status quo by establishing a dialogue on advanced spatial analytics in an academic setting. The course assumes that space is a living organism that needs to be monitored—it conceives of the building and its operations, infrastructures, and users as sources of data connected to other terrains. The projects ranged from the measurement of anxiety through vibration and occupancy studies, to the quantification of social interactions via wifi usage and sound monitoring; from the study of productivity in architecture studios via computer vision and volume of file types generated, to the exploration of correlations between digital and physical waste via software usage and occupancy analysis. 

Scales of feedback: architecture and AI

An opportunity exists in the utilization of space analytics and environmental data streams to create a closed feedback loop within the architectural design process at large. Feedback-loops between users and designers have already been deployed in other industries, where making modifications to a product is not as slow or costly as it is in architecture. Here we are referring to software development’s A/B testing: the process in which a user’s experience of two variants of the same web product is monitored in order to gain insights about performance, learn about new possible features, and test future variations.

While this closed-loop system of feedback is not currently implemented in the AEC industry, it echoes the ambitions of Nicholas Negroponte’s project “SEEK”: an evolutionary machine that learns about learning architecture. SEEK, which consisted of a Plexiglas encased, closed environment full of metallic cubed building blocks and inhabited by gerbils, was a prototypical model of an intelligent system that could monitor, evaluate and react according to its inhabitants activities. Based on the change occurring within the environment and the gerbil’s behavioral patterns within it, a computer-controlled robotic arm would modify and re-formalize the cubes’ spatial configuration. With the deployment of intelligent adaptive designs that integrate artificial intelligence and connected devices, this (perhaps rhetorical) project of the 1960s might actually be manifested in the near future.Physical computing technology is already leveraging AI to analyze the data produced by devices.

Physical computing technology is already leveraging AI to analyze the data produced by devices. Similar to Negroponte’s SEEK, products are able to establish ideal user settings from the living patterns of people. Much of the hype and fan-fare around embedded computing and predictive analytics tends to focus around metrics geared towards User Experience. This technology currently exists at specific and discrete scales—a well-known example is the Nest thermostat, which is able to set the ideal temperature within a room based on previous patterns of use. While there is of course value to this technology at the scale of the household device—location tracking to realize optimal furniture layouts, or temperature sensors for optimal insulating material in dry-wall construction—the real disruptive potential of this technology within the AEC industry exists at multiple scales of operation.

We see an opportunity for this to manifest at the scale of objects, components, and systems, which would entail the reconceptualization of the design process as part of a larger ecosystem—in which the final “architecture” would not conclude a linear sequence of steps, but rather be just one possible manifestation following the data captured through the quantification of space. There is also the opportunity to leverage this technology at the infrastructural or whole building scale. MIT’s Senseable City Lab “Underworld” project is a proof of concept to gather real-time data on the microbiome in our sewage systems. Real-time feedback at this scale could provide intelligent systems of disease surveillance, enhance efforts to predict the spread of pandemics, provide crucial data to update obsolete networks of pipes, and ultimately, according to Underworld, advance a new form of population census that cross-references biomarkers of diseases, such as obesity and diabetes, with demographic data. Indeed, the notion of feedback is not limited to the software industry, and these examples show how it can be applied at multiple scales and highlights how the quantification of space can enable an evidenced-based, adaptive and generative approach to design and form.

Measuring Representation

New (expanded) landscapes require new methods of readability. A new awareness of the built environment and the larger dynamics that it impacts will require a new mode of representation to communicate the relationships between discursive and non-discursive elements. We are in need of a visualization system that facilitates the communication of the complex correlations between physical space and the measured dimension. The N-Dimensional drawing may in fact be just that, as it merges complex dimensionality What would it mean if architects and their clients viewed the value of their real estate data on par with the actual value of real estate?with space. By dimensionality we mean all possible metrics that an architect can extract from his or her design and built environment—dimensions additional to the canonical 3+1 geometrical and temporal ones that designers are mostly familiar with.

As a tool, the N-D drawing is both descriptive and exploratory: it can be used to convey spatial relationships to different types of audiences or it can be used as a learning tool to reveal hidden relationships between physical artifacts and user behavior. Although the N-D drawing is still at its embryonic stage, this mode of representation has great potential. The architect’s role has always been to synthesize or translate complex systems into readable and legible drawings. We are confident that the N-D drawing will become increasingly easier to understand and read as it is integrated into the drawing set that students and practitioners deliver when communicating a project.

Buildings will talk. Then what?

The increasing value of data in our society goes without saying. As more and more economic assets are composed of bits, not atoms, the so-called data-fication of every aspect of our lives is increasingly securing information as the new currency. Who this value benefits should be questioned, but rather than demonize the rapid concentration of architects will have to intervene and address this process of digitization by being an actor in the shaping of new landscapes, theories and techniques for an expanded awareness of space.information and wealth, we would rather speculate on what value data can offer architecture, and whether this might lead to an expanded agency in terms of how the designer operates.

What would it mean if architects and their clients viewed the value of their real estate data on par with the actual value of real estate? How might designers leverage data as a persistent asset that can be used across an array of projects? And what would it mean to design in tandem with the quantification of architectural gestures and their effect on spatial experience? In a not so distant era, where buildings will communicate information of their own, architects will have to intervene and address this process of digitization by being an actor in the shaping of new landscapes, theories and techniques for an expanded awareness of space.

This new awareness requires not only a critical approach to how we mine data from space, but also a reconceptualization of the decision making and design iteration process. Where decision making becomes informed by a new type of evidence, and design iteration will be a non-linear perpetually adapting entity driven by a continuous feedback loop that feeds off of the quantification of space.

*Ben Hillier, The Reasoning Art: or, The Need for an Analytical Theory of Architecture