BlackBox?
The principal contributors to the BlackBox experiment:
- Sonal Beri
- Will Corcoran
- Greg Derrico
- Josh Ingram
- Zhya Jacobs
- Thomas Kearns
- Neil Katz
- Justin Nardone
- Doug Pancoast
- Bernard Peng
- Joel Putnam
- Matthew Shaxted
- Heechan Shin
- Michelle Swanson
- Andy VanMater
- Ning Wang
- Wei Wang
- Britt Woolf
I wish I had a share in a hot startup for every time I have been asked, "So...what is the BlackBox?" Students...people from other firms...friends...family...even other SOMers. To be honest, even I found it difficult to convey in a just few words, and the words I tended to use were probably too abstract and vague to get the idea across. Perhaps the concept itself was somewhat ill-defined, but what was always most effective was to simply present the work; no matter the audience, the stories about the work never failed to make it immediately clear why the group existed and what made it unique. I wish I also had a share for every time someone exclaimed afterwards, "I can't believe this is going on at SOM!"
As I write this in early 2015, I have had some time and space to reflect back on my time as the founder and director of the group, as well as the evolution and significance of the work itself. As a means of preserving these thoughts while they are still fresh, and to establish a partial record of the work, I created this website. The vast majority of the group's work was authored not by me, but by the remarkable stream of people I was able to bring into the group. Part of the challenge of identifying a cohesive narrative for the story is that to a large extent, the work that was generated was determined by the unique skills and interests of each individual; BlackBox people tended to have many talents and many interests.
Origins
The story of BlackBox began in 2005 when I decided to take a mid-career sabbatical from practice, resigning my position as the Design Director for an architecture firm in South Carolina, and returning to school full-time to pursue a unique Master of Engineering degree at the Stevens Institute of Technology in Hoboken, NJ. The program at Stevens was a brand new venture, founded by Harvard GSD alum, John Nastasi, and was called the Product Architecture and Engineering (PAE) Lab. The essence of the PAE program was a deep and semi-structured immersion in the rapidly emerging intersection of design and computation. The curriculum included parametric modeling, computer programming, computational simulation & analytics, numerical optimization, digital fabrication, UX design, sensing & actuation, and other concepts that were on the frontiers of application in architecture and other disciplines of design.
While I was enrolled in the PAE program, the New York office of SOM started sponsoring research, and I was the leader of the first student team (along with Josh Ingram and Charlie Portelli) that was engaged to work with the SOM design teams, applying our nascent computational skills to "live" projects in the office, and enlightening the SOM staff to the powers of these skills. Some of that early, groundbreaking work from those engagements is documented on this website.
At the conclusion of the PAE program in the Spring of 2007, I accepted an offer to set up an elite group in the Chicago office of SOM, modeled largely after the concept that we had defined with the NY office as students. So I arrived in Chicago that May, along with Josh Ingram, Will Corcoran (both classmates of mine at Stevens) and Britt Woolf; and BlackBox was born. The mission of BlackBox was - simply and broadly - to be the catalyst for integrating more advanced forms of computation into the design processes of the office. The name was appropriated from the fields of computer science and engineering where the term refers to "a system, device, or object (or algorithm) which can be viewed in terms of its inputs and outputs, without any knowledge of its internal workings." (Wikipedia) The expectation, in my mind, was that one of the primary products of the group would be algorithmic "black boxes" - design tools to be used by the design teams. Plus, the name just sounded kind of cool and evoked a useful amount of intrigue.
Josh, Will, and Britt were all trained as architects, but their design talents were supplemented with unique computational skills, and more importantly, a fervent interest and faith in the ability of computation to take design in directions that were not otherwise possible. I would say that this was a defining character trait of all the people who were associated with the group - a regard for computation not as a threat to the design process, or even as just an interesting novelty, but as a vital extension of the designer's psyche.
Up and Running
The process of integrating effectively into the design work of the office - and the consciousness of the designers - took a sustained and proactive effort. No such group had existed in the office before. Angst over internal business accounting (chargeability) structures led to our being assigned to the IT department, which was a double edged sword: on the one hand, the pressure of chargeability was somewhat alleviated since IT was fully accounted for as overhead; on the other hand, the culture of the design studios was that designers rarely (if ever) thought of bringing IT people into conceptual design conversations.
So the initial stretch was spent forging relationships with the designers, introducing them to the relevance and capabilities of our group, getting them to consider making adjustments to their well-entrenched design processes, and later, getting project managers to allocate time in the project plans for "BlackBox hours" (I wish I also had a startup share for every time I got a call or email from a project manager wanting to know why there were BlackBox hours being charged to their project!).
In the beginning we found it easiest to engage with the structural engineers, many of whom were similarly versed in coding and algorithmic thinking, and we immediately started putting the Genetic Algorithm to work on a number of structural optimization problems and research initiatives. As we began to interface more with the architectural studios, we found that our most valued skills were the parametric modeling of complex geometry, solar shading and exposure analysis of conceptual designs, and the use of scripting to automate and expedite tedious tasks. We formed a highly productive bond with one of the design studios in particular, mainly because of the fertile curiosity of the senior designers about our capabilities, and their eagerness to explore the possibilities.
Tool-making was a fundamental part of the BlackBox culture. Often times, the project engagements we took on required the simulation, measurement, or optimization of some aspect of the project. If no such tools existed to carry out these operations, we would routinely custom code our own. Some of these tools were relatively simple, but sometimes they were quite ambitious, requiring many weeks of coding. As pressures mounted to maintain very high rates of chargeability, it became more and more difficult to develop the more ambitious tools, especially if they required more time to develop than a single design project could support. Even when it was clear that the tool could provide critical benefits and efficiencies on many projects going forward, there was not sufficient support to execute the work as an overhead initiative or to even spread the cost across multiple active projects.
Pivoting with the Recession
By the time the recession hit the office around 2009, BlackBox had grown to eight people. As the cost-cutting measures became brutal reality, the non-chargeable activities within the group were an immediate and understandable casualty. I was even braced for a total dismissal of the group, but instead the strategic decision was made by the Partners to move all the BlackBox people into the design studios in order to maximize chargeability. Bullet dodged (although I never did fully exhale).
There was both positive and negative fallout from this strategic shift. On the positive side, in many cases the design projects benefitted from a more direct involvement of the BlackBox people and the application of their skills. The skills and knowledge became more effectively distributed, as opposed to residing in a centralized specialty team. The removal of this perceived barrier was definitely a good thing. So was the fact that a lot of good work was done that had immediate impact on design quality and project productivity.
There were also negative consequences, however. In many cases, the studio heads were ill-equipped to understand and exploit the skills and talents of the people they had inherited. Project delivery pressures often necessitated that some very talented people were relegated to some very menial tasks. Of course this happens all the time on a project when it is time for everyone to just roll up their sleeves, but when it continues for a significant amount of time, the damage to morale becomes an issue. Predictably, we lost some extremely bright people when they were no longer working on challenges they found interesting.
An additional consequence of breaking up the group was that it became even more difficult to advance any research or long-term development initiatives. To counter this, I began to cultivate relationships outside the office in order to shift some of the load. I became much more involved with the architecture programs in the local universities, with Argonne National Lab, and with other large corporations. With the help of Design Partner, Ross Wimer, the design studios at IIT became a particularly good testing ground for various initiatives. We wrote numerous grant proposals in partnership with various academic institutions (with varying degrees of success) and collaborated on important research work with companies like Dow Corning, Permasteelisa, Guardian, and IBM.
The Evolution of the Work
The work coming out of BlackBox evolved considerably over the first 7+ years of the group's existence:
- Parametric modeling was always a constant within BlackBox, however, Grasshopper emerged during that period and made parametric modeling accessible to the rank and file in the design studios, at least for conceptual stage studies. Digital Project/ Catia was still a defining tool of BlackBox people, but its use shifted from conceptualization to become the go-to platform for complex technical geometry because of its superior ability to explicitly control surface geometry and to manage very large geometry files.
- There were a number of tower projects generated in this period that incorporated highly complex, doubly-curved enclosure geometry, which led to the development of considerable internal expertise in the "rationalization" of these surfaces into buildable curtainwall panels. The double curvature of these skins often led to panelization strategies that incorporated "cold-bending" techniques, and we also gained considerable expertise with respect to this fabrication/installation technique.
- During this period, we also saw the emergence of Revit as the go-to platform for project documentation and delivery. Although BlackBox was not a catalyst in the adoption of Revit, we did play a leading role in developing interoperability tools and standard workflows for moving models and data back and forth between Digital Project and Revit, as well as Rhino, Grasshopper, and even the structural analysis/modeling software tools that the engineers used.
- During the first half of this period the focus was almost exclusively on architectural projects, but I then directed a turn toward our urban design studio. Having not spent much time in this discipline, I was somewhat surprised to discover that there were so few urban design oriented computational tools available to the urban designers. As a result, the typical urban design process was largely driven by the intuitive experience of the designers, with very little feedback in the way of hard metrics or analytics. As IT companies like IBM and Cisco began to promote the concept of the "Smart City", urban designers were largely marginalized in the conversation because of the fact that they had spent so little time thinking about the relationship between digital technologies and the design of the city. Therefore, I initiated the conversation within the SOM urban design studio and leveraged BlackBox to begin developing better digital tools geared specifically for urban design processes.
- The work on cities led to a greater awareness of, and interest in, the concept of "Big Data" - another topic that was seemingly capturing the imagination of the outside world while the design firms were left to wonder what it had to do with their work. Through BlackBox, we initiated the internal conversations about the definition of big data, its manifestations in an architecture, interior design, urban design, and engineering practice, how to exploit our particular data sets, and how to collect and mine data that we weren't already collecting, both for Business Intelligence, as well as Design Intelligence.
- The shift to working with larger volumes of data precipitated the exploration of a technical paradigm shift involving yet another buzz concept - cloud computing. With some assistance from computer scientists at Argonne National Lab, we developed workflows designed specifically for execution on clusters of multiple networked computers in parallel, as opposed to the status quo of leveraging individual workstations. With cloud computing we were basically able to increase the number of simulations we could run by a factor of the number of cores that were networked in the cluster. We used this computing power to set up "parametric sweeps" in which we defined a large number of design variables and then ran simulations of every possible permutation of the design. In some cases this parametric design space of possible designs ran into the millions. Depending on the amount of time took to run each simulation and the number of processors in the cluster, we could have complete results by the next morning, and know which design within a population of 15,000 possibilities, performed the best with respect to energy consumption.
- Another wave moving through the software development world was open source web-based technology, and BlackBox was actively exploring the application of these technologies to our work at SOM. These explorations were initiated largely because of the interests and expertise of a couple specific individuals, and they allowed us to take a critical look at the value propositions of when web-based open source solutions made sense versus desktop based applications that leveraged the APIs of commercial software programs.
About this Website
The three main sections of this website are the Thematic Overviews, the Project Index, and the Essays + Talks. The Thematic Overviews are a collection of thought pieces that look at the work of BlackBox in the context of a number of computational design themes that were relevant within the industry at the time (parametrics, simulation, optimization, big data, etc.), attempting to illustrate the themes and relate them to an architectural practice through the BlackBox projects.
The Project Index is a straightforward archive of some of the work that BlackBox people produced, some of which was generated while at the Stevens Institute of Technology as students, before the BlackBox group was formally established. The work ranges from abstract speculative explorations to the highly specific development and application of digital tools on architectural and urban design projects in the office.
The Essays + Talks section contains a number of published articles I wrote during my time with BlackBox, as well as a few recorded talks that I made at various occasions. The final section is a brief personal resume.