Isthmus’s Boudewijn Thomas on Parametric Urbanism
This article has been reshared from isthmus.co.nz
For the month of May we’re putting a focus on Design &…Business with featured thought leadership articles, events & workshops, and design projects from and for our community. You can check out more from the Design & Business series here
Boudewijn Thomas trained as an architect in The Netherlands, before spending seven years with MVRDV. At Isthmus Boudewijn has been leading a team that implements state-of-the-art digital design tools including parametric design, analytics and data-driven methodologies. Here he talks about the opportunities and challenges of data-driven design through three Isthmus projects he has worked on.
With the emergence of computational design, it is now possible to quickly generate thousands of design options, enabling designers and stakeholders to test ideas at much higher speeds and with more rigour than ever before. Designers are starting to focus less on the creation of single drawings through traditional CAD software and are shifting to designing with their own algorithms. These so-called parametric models contain sets of rules, conditions, and inputs to generate many drawings simultaneously, often accompanied by data. Together with stakeholders, the idea is that these models can be explored in real-time, leading to more engaging and valuable interactions. At Isthmus, we are increasingly leveraging these parametric models in our project work, ranging from design planning to urban design and early architectural stages such as business case modelling and feasibility.
Plan Change 33 study, Tauranga.
Parametric models give us more flexibility and increase the legibility of our process, improving collective decision-making. In our work on the Plan Change 33 urban intensification measures for the Mount Maunganui area in Tauranga, we created a parametric model to test the proposed landscape measures and visualize the impact on the future urban fabric. Not only did this enable us to do a better job as designers and planners, but it also made the information more legible to the people who live and work in Tauranga increasing public buy-in. In the model we control parameters such as the maintaining of views to Mauao and responding to a beautiful coastline and, perhaps most importantly, how all the different moves interact with one another in 3D. Through a set of steps, a 3D model was generated indicating volumetric ‘envelopes’ that represented potential development areas that can be explored from atop and from eye level.
On project like this, priorities regularly shift and new data comes to light during the development process. This often significantly disrupts production workflows and results in lengthy revision times. Complex documents such as these may turn into text-heavy and difficult-to-digest reports which makes understanding them even more challenging, limiting our collective ability to make well-informed decisions and potentially leading to all kinds of issues down the line. By using parametric models in our PC33 study, we were much better positioned to implement and understand proposed changes—traditional techniques which would have required days of rework, but instead we were able to update the model in a matter of hours, allowing for more iteration and improved project outcomes.
Te Ara Tupua Coastal Infrastructure
Data-driven design is a fundamentally different approach, requiring us to think in new ways and use other means to interact with our designs. Traditionally we would draft a couple of options, pin them on a wall, compare them, and take our learnings into the next round. If you made the right assumptions, you would look to get a good result, otherwise, it’s back to the drawing board. Iterations are limited, so we rely heavily on skilled designers to make the right calls to control costs. Generating many options doesn’t necessarily solve this challenge. Faulty models and naïve algorithms can derail processes quickly, cloaked under unwarranted trust in data and ill-transparent processes leaving inexperienced designers out to hang high and dry. A well-rounded, data-driven approach therefore aims to bring the information back to a digestible size, explaining what goes in, how it is processed, and what comes out. We need to understand what story the data is telling us and how that data is framed. By being transparent, the process is accessible to a wider audience allowing considerations and expertise of the group to be more likely included in future iterations, ultimately improving the quality of outcomes.
As part of our design role for the Te Ara Tupua Alliance, a coastal resilience project on the edge of the Wellington harbour, we helped to test design options for the revetment walls that will be constructed out of interlocking, prefabricated concrete Xbloc units. The Alliance’s vision for this element included ecological and cultural objectives whilst keeping track of engineering and construction parameters. We wanted the design to feel bespoke and natural, as if the rocks have been there for many years, shaped by the ocean they sit beside. A certain level of randomness was required to achieve this outcome, different types of blocks, different colours, and different amounts of each of them. We created a script that would help us test proportions and arrangements. The script allowed us to set out the Xblocs and create different sets – one block, two blocks, four blocks, eight blocks, etc. until we arrived at the right balance. We added a shuffle parameter to the script to create randomness. Finally, we added an automated QA that would highlight and help reconcile specific block combinations that were not allowed from a technical point of view. Using this script, we were able to visualise options, analyse cost and ensure technical feasibility—all at the same time. We were able to further refine the script by adding new parameters such as the ecological blocks below the water line and refine earlier parameters based on new costing assumptions and engineering inputs.
Urban intensification studies, Tāmaki Makaurau.
Holistic urban design requires input from the widest possible audience, meaning clarity on the process and the outcomes is equally important. In a recently completed study, we used parametric modelling and data-driven design to test and understand high-level future density scenarios for several suburban areas. 3D models were developed in collaboration with a business case team that fed data back into business case modelling on their end. This allowed the team to verify that the numbers added up to support favourable urban outcomes while making a sound business case for proposed transport infrastructure.
The urban models needed to react to several criteria including property rights, financial data, lot shapes and sizes, and numerous regulations that would limit heights and configurations of preferred strategies. We challenged these constraints to investigate future development scenarios, merging lots, intentionally ignoring regulations, and continuing to adjust the parameters of our model. Results ranged from conservative pop-up development to opportunistic mixed-use scenarios that could even be set out in time to inform long-term development strategies. By combining 3D visuals with data analysis, we were able to create more appropriate holistic responses by asking the right questions and bringing these back into our models. Is a maximum development scenario realistic? What “feels” right? Do we need to be greener? How do we deal with parking? Will that house still have a view? How walkable is the neighbourhood? What is the impact on views and sunlight?
Parametric Urbanism.
We want to create the best outcomes for our land, our people and our culture. This means improving the places that people live, work and play, requiring—amongst other things—densification of our urban environments. But densification is no easy task, there are many factors to consider that require involvement from experts and the people that already live there. Fortunately, as designers, we have more tools at our disposal than ever before to tackle this. By combining 3D models with data, we can understand, communicate, and develop spatial outcomes better. Having the numbers to support the feasibility, we can improve collaboration and involvement of expert and non-expert groups and ultimately unlock the latent potential in our urban environments.
See more from the Design & Business series:
