A+u magazine was recently granted an exclusive interview with the co-founders of Flux, the Google[x] startup whose mission is to harness data to automate architectural and urban design. The discussion is one of 14 essays and interviews from leading urban technologists in the current November issue, Data-Driven Cities.
“We began our exploration with the premise that buildings and the sustainability of our modern lifestyle are deeply intertwined. In addition, buildings – more specifically, housing – is an issue of human dignity. We wanted to find ways to apply Google-scale thinking to tackle these important issues," says co-founder Nicholas Chim in the interview.
Read on after the break for a+u magazine's full interview with Flux co-founders, Nicholas Chim and Michelle Kaufmann. And check out the November issue of a+u magazine, available in digital and print editions, which features new essays by Carlo Ratti (MIT), Dan Hill (City of Sound), Alastair Parvin (Wikihouse) and more.
a+u: Flux was born out of Google X, Google’s moonshot1 factory, dedicated to finding radical solutions to huge problems. So, what kind of world-changing solutions is Flux working on?
Nicholas Chim (NC): We began our exploration with the premise that buildings and the sustainability of our modern lifestyle are deeply intertwined. In addition, buildings – more specifically, housing – is an issue of human dignity. We wanted to find ways to apply Google-scale thinking to tackle these important issues. We started with nearly a blank slate, literally because the software engineers on the team had no prior background in architecture or construction, then engaged thought leaders, interviewed practitioners, built prototypes, and tested our thinking with a broad cross section of industry. Through this process, we discovered a strong desire to address the industry’s challenges, yet practitioners are caught in the realities of today’s business relationships, legal structures, risk tolerance, and design tools. To break this cycle, we are focusing our efforts on improving collaboration during planning and early design, enabling data-driven decision making, reducing information latency, and building knowledge communities.
a+u: Is Flux inventing new BIM (Building Information Modeling) software?
NC: If “BIM” is defined as the 3D representational model of the building, then we are not building new BIM software. BIM is a mature technology; design and construction firms have invested heavily in it to achieve greater efficiencies and tackle increasingly complex projects. Instead, we’ll focus on integrating our system with industry-leading BIM and CAD platforms.
We are building two classes of tools: the first class connects existing tools together to allow seamless execution of complex workflows, and the second class captures design intent. For instance, the rules to lay out exit stairs are fairly easy to explain. When converted to software, our system can apply those rules to generate a new design based on a project’s unique requirements. Moving design logic into software promotes reuse, knowledge sharing, and continuous improvement. This is the foundational concept needed for our industry to achieve scale.
Collaboration is also a key requirement of our system. Collaboration around a shared model allows each expert to maintain their specialized domain model, while providing stakeholders with a holistic view of the entire project during the early design stages. Our tools provide decision support by computing key metrics such as construction cost and life-cycle operating cost in near real-time.
a+u: The first software you’ll release harnesses public urban data to inform development at its earliest stages. Can you explain how it’s used and who it’s for?
NC: Our first product will be Flux Metro. Its immediate utility (as of this interview) is limited to Austin, Texas, but it’s an experiment. Two key observations surfaced when we were researching urban real estate development: that the real estate entitlement process is laborious and expensive, and that site context is the primary driver in building design. We noticed that real estate developers, land-use specialists, and architects were spending considerable time gathering and consolidating data from a multitude of sources to understand development potential and constraints. Furthermore, this process is repeated for every candidate site by every interested developer.
Our first inclination is to help them integrate and manage these data, but that is just a starting point. We needed to create a 3D experience where they can visualize and comprehend the amalgamated data. Furthermore, we needed to show the cityscape not only as it is today, but as it might be as planned, so we painstakingly digitized Austin’s planning codes. Our product can interpret the spatial relationships that form the basis of zoning codes including district overlays, floodplains, distance from a specific street, parcel adjacencies, corner lots, and Austin-specific view corridors.
Most of the data in the product can be accessed over the web for free after registration at https://flux.io/metro. We do charge a fee per land parcel if you want a more in-depth interpretation of the planning codes, which replaces several hours of combing through dense zoning and land-use codes. Once we prove out the model in Austin, we’ll expand Flux Metro to include more cities.
a+u: Seems it’s not only a tool for architects and clients to easily gauge feasibility. It could also be very valuable to city governments.
NC: We were pleasantly surprised when we showed early prototypes of Flux Metro to city planning officials. Many of them have internal initiatives to harmonize datasets and provide access to the public, but progress has been slow due to lack of budget and expertise. Flux Metro gives planners a three-dimensional view of the zoning codes as they’ve never seen before. We showed them parcels that had so many competing zoning overlays that building anything short of a car park was impractical.
Beyond visualization, planners immediately saw the benefits of using Flux Metro to plan and communicate changes to the zoning codes. They want to see the difference between the current codes and the proposed codes, and to be able to share proposals with community stakeholders to get feedback and build consensus. Providing this level of transparency and clarity will hopefully allow cities to strike the balance between the needs of existing communities, improving the urban living experience, and economic development.
a+u: So this is the first demonstration of a larger vision concerning data and cities. Can you tell us how this fits into your bigger picture?
NC: The key challenge for the industry is to achieve economies of scale. There is no shortage of innovative ideas, inspirational designs, and project leadership, yet their impacts are localized to a very small percentage of projects. How can our industry increase the rate of innovation and improve the quality of design for the mass market? We need to disseminate ideas, cull the bad ones, improve on the good ones, and encourage the generation of more ideas. But ideas alone are not enough, the tools to actuate the ideas need to be made accessible and evolve.
Naturally, we should take advantage of low-cost cloud computing, mobile devices, and ubiquitous data networks. We also need to rethink the way we add value to our clients, the building owners. We have to focus beyond construction cost. Instead we have to deliver infrastructure with greater agility, while paying closer attention to human health and environmental impacts. The full-scale design problem is far more complex than our current scope.
It will take many years for this vision to fully materialize. We’ll need to try out many ideas, some of which will not work. During our early development, we’ve been fortunate enough to enlist a committed group of partner firms. But in order to scale our efforts, we will need help from all corners of the industry.
a+u: How do you predict the proliferation of urban data will transform the role of architects?
Michelle Kaufmann: We become architects because we love to design and we thrive when we are creating environments that are inspiring, beautiful and sustainable. However, we end up spending most of our time doing mundane work such as data collection, code reviews, counting, e-mails, and repetitive work that has previously been done. I don’t know of an architect who loves reading code books and working on area calculations for multiple constantly changing schemes. Yet this is critical work and its accuracy is essential to a project’s success.
By having urban data and code requirements provided instantly, it empowers architects to study many scenarios with the knowledge of how it aligns with city planning. We can be more intuitive (which most architects prefer when designing) with results that are more contextual. This allows the computer to do what it does best while allowing the architects more time to do what they do best which is innovation, creativity, and making buildings that improve our communities and peoples’ lives.
Having tools like Flux Metro can also help architects’ business models, especially for firms that are moving towards compensation models based on the full value created, rather than on hours spent. Data-driven collaborative cloud-based technology helps with this new business model by allowing us to design better buildings in less time.
NC: Context is a primary consideration in architectural design. By creating a data-rich virtual environment, our tools can help architects assess the fitness of a proposed design in relation to its context. This contextual awareness is especially valuable when performing energy modeling, daylighting analysis, and traffic simulation.
a+u: The software industry has appropriated the term architecture (for example, software architects, information architecture, etc). Do you find strong connections between these two very different industries?
NC: Software engineering has indeed borrowed many concepts from the world of architecture including co-opting the title, “architect”. The structure of software with its frameworks, scaffolding, and interconnects parallels the structure of buildings. Christopher Alexander’s seminal work2 on architectural design patterns inspired the formalization of design patterns in software. Like building design, software development has a strong bespoke design culture; software projects are just as likely to run over budget and schedule as construction projects. Of course, the obvious difference with software is that copies can be made essentially for free.
The software industry has been evolving at an exponential rate since its inception three decades ago. How can software engineering return the favor to architecture and help improve the scale economies of our industry? How can we amplify innovations in order to meet the twin challenges of urbanization and sustainability? Seeking answers to these questions motivates our team and our work.