Information in the Age of Big Data
As designers, our relationship with information in the big data age is at once both empowering and potentially paralyzing. From mapping data, to environmental analyses, programmatic data, and social media analysis, there are myriad inputs that we could use to drive our designs. What’s more, as our tools continue to advance and our software gets smarter, the ways in which we are able to use this data will unlock more and more possibilities.
By Michael Lee
Date
July 16, 2020
Information is everything.
It’s a common sentiment, but I don’t think it has ever been spoken more loudly than it was during the onset of the COVID-19 crisis. In the early days of the lockdown, I, like most people, was following the daily case numbers and projections with the hopeful optimism that the pandemic would be over relatively quickly. Then when it wasn’t, I, like most people, continued to watch as the uncertainty unfurled with our collective lack of understanding of case numbers; this was due to lack of testing, the ramifications of asymptomatic disease transmission, the extent of community spread, and more. It seemed there were a lot of holes in our understanding of the situation and we only ever had half the picture. Information, evidently, was everything.
As designers, our relationship with information in the big data age is at once both empowering and potentially paralyzing.
There are seemingly endless amounts of data that, when we build into our design models, can seem overwhelming regarding where to start. From mapping data, to environmental analyses, programmatic data, and social media analysis, there are myriad inputs that we could use to drive our designs. What’s more, as our tools continue to advance and our software gets smarter, the ways in which we are able to use this data will unlock more and more possibilities.
Using tools like parametric and generative modeling, we are able to build customizable algorithms that can use any or all of these data points as inputs. We are then able to generate and test thousands of different options for any given design problem to see what configurations achieve the best balance of performance factors. What’s more, as these methods become more ubiquitous, the ecosystem for these tools continues to get more powerful with each passing day as new research and plug-ins become available that expand the capabilities of the platform. This is often a scary prospect for designers, one that sounds very much like the machines are taking over, however I’d argue that this could not be further from the truth.
One really important factor that is not to be lost is that these data points can only capture a part of the picture.
The design of spaces is a holistic practice. One that balances the art of creating experiential spaces with the technical sciences. While we can rationalize our designs with these data points to optimize their indices for success, it doesn’t mean that these are the only measures of a good design.
Computers are brilliantly useful tools to crunch numbers. They are, however, not inherently smart tools. They are as intelligent as we program them to be, and smart insofar as we can quantify aspects of design. Certain performative aspects like square-footage efficiency and solar analyses are directly measurable and relatively simple to use to drive a design. But projects that can be simplified to just these measurable aspects are few. Trying to quantify more subjective aspects like aesthetic character or human-scale comfort can perhaps be measured by proxy, but the way we generalize these aspects to be quantified I think will always invite contention and won’t capture the true value judgements designers regularly make (for now I should say, as artificial intelligence research marches ever forward).
To successfully integrate the power of computation into the very often complex and ambiguous design process, we must approach it with full understanding that technology in design is a tool to help us.
There are plenty of tools to help us analyze and make connections, and I have no doubt that using computational methods in design can almost always benefit a design process. However, the extent to which it is applied is a critical consideration. At the end of the day, humans will be the interpreters of design, not computers, and the decision of where to stand on the spectrum of “data-driven” to “data-informed” is up to the designers to be the interpreters of design problems and play the final curatorial role.
Michael is a computational designer holding a Master’s Degree in Architecture from the University of Toronto. Since joining IBI, Michael has brought innovative tech-driven design solutions across the firm’s wide-ranging practice areas including architecture, interior design, and urban planning. As Regional Computational Design Lead, Michael brings his expertise in parametric modeling, analysis, and design automation to improve workflow efficiency and design capability through all aspects of design.
Written by Michael Lee
Associate | Regional Computational Design Lead
Toronto East, ON
