Beyond Surfaces: Applying Intrinsic Geometry Processing in Art and Computational Design
Interview with Math Whittaker, New Balance
Math Whittaker’s work is the perfect example of how a computational mindset enables designers to move fluidly across applications and scales, whether developing wearable technologies, exploring digital art, or advancing performance footwear.
As a Computational Design Programmer at New Balance, his approach is rooted in intrinsic geometry and abstract rule-based thinking, allowing design intent to emerge from the underlying behavior of surfaces and structures.
In this interview, Math reflects on how working with code, not just existing software tools, has allowed him to shift between scales and applications, how his current commercial work connects back to early experiments in generative art and physical computing.
As he prepares to present at CDFAM Amsterdam, he shares insights into how ‘intrinsic geometry processing’ can unlock new aesthetic and functional possibilities beyond conventional 3D modeling.
Can you start by telling us about your role at New Balance? How did you first get involved in computational design, and how does your current work connect back to your earlier experiences in art and wearable technology?
My role title at New Balance is Computational Design Programmer II, and I work within the computational design department, which is part of the innovation design group.
My responsibilities are quite varied. On one hand, I focus on developing custom workflows and tools that aid in the design and development of our footwear products. On the other, there’s a strong research side to my role where I’m constantly looking for new techniques and technologies that we can integrate into our design process. This encompasses both aesthetic considerations and exploring new methods to study data and use that information to inform the design process.
My first exposure to computational design, specifically using Grasshopper and Rhino, was around 2010, suggested to me by an architect at Zaha Hadid.
While I was always interested, it was at Manchester School of Art in 2013 that I truly committed to the computational design process. This was partly out of necessity, as traditional design skills like drawing are not my strength. However, I also became genuinely fascinated by how computational design could enable the creation of subtly complex designs, and how this process so easily allowed us to input real-world data, either to inform the design directly or to create specific visuals.
In terms of how my current work connects back to my earlier experiences in art and wearable technology, I’ve always found it slightly challenging to see a stark difference between various design disciplines from a computational design perspective.
Of course, there’s specific knowledge unique to each field, but from a computational design standpoint, I feel the fundamental process remains consistent, and ultimately, the underlying mathematics doesn’t change. If anything, I find it quite interesting how much similarity there is. Across all three areas – art, wearable technology, and now footwear – I’ve often ended up looking into the same algorithms.
The key difference lies in the application and how these processes are displayed or rendered, whether digitally or physically. To be honest, what I most enjoy about computational design is thinking about design in the abstract, where you’re conceptualizing the design’s method through rules, and then allowing that design to develop itself through digital simulations.
New Balance has had a strong presence at past CDFAM events through Onur Yüce Gün’s presentations. How does your work relate to or extend what Onur has shared previously and how do you approach problems or design challenges together, or from distinct angles?
The differences and similarities between Onur’s presentations and mine extend to how our computational design department operates at New Balance.
Onur’s previous presentations at CDFAM offer a broader, more encompassing view of how machine learning and computational design are integrated within the design world. This contrasts with my presentation, which focuses on a much more specific area of computational design and its practical applications.
This dynamic closely mirrors the structure of our computational design department at New Balance. Onur, as the lead, provides strategic oversight and defines the team’s direction. Him coming from a technical background help us a lot while we align strategy with individual applications.
My colleagues Dana Maringo, Louis Leblanc, Sam Whitworth, and I each cultivate specific areas of computational design expertise. This specialization broadens the range of problems our department can effectively address.
Through this collaborative structure, Onur can seek our specialized advice on specific design challenges. In turn, we consult with Onur on the optimal implementation of our solutions, ensuring a cohesive and effective approach to problem-solving.
From a software perspective, what does your computational workflow look like today? What commercial software or platforms are you using, and where have you found it necessary to develop your own custom tools?
When it comes to commercial design packages, I primarily rely on Rhino/Grasshopper. However, these tools make up only about 10-15% of my daily work, typically for quick code sketching and idea testing. My main focus is on direct programming, primarily in Javascript, C++, C#, and Python, in that order.
There are several reasons for this. While tools like Grasshopper and Blender are brilliant, their steep learning curve can be a barrier for designers.
For computational design processes that are ubiquitous across our products at New Balance, I often develop small web applications using Javascript, HTML, and CSS. These are designed for ease of use, ensuring broader access to the benefits of computational design within the company.
Beyond accessibility, I deeply value the freedom that direct coding offers. It allows me to create custom interfaces and modify algorithms far more easily than I can with Grasshopper.
Crucially, there are also significant performance benefits to developing specific design tools through code. While Grasshopper is an incredibly powerful tool, it does have limitations—for instance, when handling large datasets or running complex algorithms on 3D geometry, it can struggle, especially with operations that require advanced conditionals and loops. This is where the power and efficiency of C++ development become invaluable.
Where you see gaps in the existing software landscape, particularly for designers working with geometry at the level you’re exploring, what kinds of tools or capabilities would you like to see developed to support more advanced or expressive computational workflows?
That’s an interesting question, and my perspective might be a bit different. One thing I’ve really come to appreciate from writing small, specific software and apps for design is just how light and computationally efficient they are. They simply do the job they’re designed for, without any unnecessary overhead.
I believe any software should aim to be exceptionally good at something specific rather than trying to do everything. For instance, while it might seem ideal for Grasshopper/Rhino to handle volumetric meshes like tetrahedral meshes more effectively, even if it did, I’d probably still only use it for sketching out ideas. In the long run, I wouldn’t want to deal with the significant overhead that would come with those meshes being linked into Grasshopper’s much larger codebase.
In a more direct answer to your question, I think that there’s almost certainly a design software out there that can do what you want to do. It might not be a feature within the design tool you’re currently using, but it likely exists as a specialized solution. And in many ways, I see this as a positive thing. If a single design software truly did everything and did it all exceptionally well, I feel that tool might become so massive that you’d end up feeling lost in its complexity.
I believe tools having limitations is actually a good thing; it keeps them focused and, in turn, helps our workflows.
Your talk at CDFAM will focus on applying intrinsic geometry processing techniques to design and aesthetics. Tell us a little about what you will be presenting, and how do you see these methods expanding creative possibilities beyond conventional surface modeling?
My talk at CDFAM, ‘Beyond Surfaces: Applying Intrinsic Geometry Processing in Art and Design,’ will delve into methods that go beyond traditional explicit 3D modeling. Instead of focusing solely on defining a shape’s form with global coordinates, we’ll explore techniques that leverage intrinsic geometry processing properties – characteristics inherent to a surface itself, regardless of its position or orientation in 3D space.
I’ll be demonstrating the practical application of these methods through:
• The generation and application of smooth vector fields on complex meshes, illustrated with a personal jewelry design project where flow and orientation were critical.
• The use of the Cotan Laplacian as a fundamental surface operator, specifically showing how it enables Reaction-Diffusion systems to run directly on a 3D mesh, overcoming the limitations of 2D texture mapping.
• And, crucially, how we can integrate other intrinsic properties, like Shape Index, to actively modulate and control these Reaction-Diffusion patterns, making them intelligently respond to the local curvature of the design.
These techniques significantly enhance creative possibilities beyond traditional surface modeling through:
1. Form-Aware Design: Instead of designs being imposed from an external 3D space, intrinsic techniques allow patterns and features to truly grow and adapt to the object’s inherent geometry. This means patterns that naturally flow, densify, or change based on curvature, ridges, or valleys – making the design feel intrinsically integrated.
2. Seamless & Consistent Aesthetics: Patterns defined intrinsically remain consistent regardless of how the object is rotated, translated, or posed. This overcomes distortions and seams often encountered with explicit mapping, enabling truly seamless and stable surface treatments.
3. Unlocking Organic Complexity: They provide tools to generate intricate, organic, and self-organizing patterns that are incredibly difficult, if not impossible, to achieve through manual or purely explicit modeling. We’re tapping into the mathematical ‘language’ of shapes to create new aesthetic expressions.
4. Sophisticated Control: Beyond just generating complexity, we gain sophisticated control mechanisms. We can guide the orientation of discrete elements, align anisotropic textures, or even influence complex generative systems like Reaction-Diffusion based on the very character of the form, pushing design expression into exciting, novel territories.
What do you hope attendees will take away from your presentation, especially those coming from a background in industrial design or applied engineering, and what are you hoping to gain from participating in CDFAM? Are there particular kinds of conversations or collaborations you’re hoping to build within the community?
I hope my presentation illuminates an area of computational design I find incredibly powerful and deeply interesting. I aim to illustrate to industrial designers how this often academic or hidden area of mathematics can be a powerful tool for generating designs intrinsically informed by their underlying geometry, leading to more aesthetically pleasing and functionally intuitive visuals.
For applied engineers, I believe the concept of leveraging inherent geometric information to inform design will deeply resonate with their work processes. Witnessing how these properties can be applied, I hope, will inspire new computational approaches.
I want to inspire attendees to explore the ‘hidden information’ within the shapes they design. As the William Blake quote in my presentation, ‘To see a world in a grain of sand,’ concisely describes my view of intrinsic geometry, I hope it resonates and sparks a similar sense of discovery in others
Join us at CDFAM Amsterdam to connect with Math Whittaker, Onur Yüce Gün, and others working at the intersection of computation, design, and manufacturing, from footwear innovation to adaptive architectural systems.
Whether your focus is product, performance, or process, the event brings together practitioners applying computational methods across disciplines and scales.