Form Follows Force: Data-Responsive Stochastic Algorithms For Additive Footwear
Interview with David Burpee
David Burpee is a computational designer based in the Pacific Northwest who has spent more than a decade developing design tools and engineering strategy for brands including Nike, PUMA, Brooks Running and General Motors.
Trained as an architect at USC, he worked on high-rise and supertall projects before moving into the ‘faster moving fields’ of footwear, consumer products and medical applications.
Ahead of his third CDFAM presentation, following talks in New York and Barcelona, he discusses Morphxgen, his new footwear brand, and how gait data and simulation can drive printed lattice structures that respond to the way people move.
This will be your third CDFAM, following computational morphogenesis in New York last year and the Sequence insole generator with Christoph Brandin in Barcelona this spring. What has changed since then, what is MorphXgen, and what will you be covering in Tokyo?
It is a bit surreal presenting at my third CDFAM, something keeps me coming back I suppose. I am still exploring biomimetic and algorithmic design for footwear, still building generative systems and emergent design tools.
I am actually starting a brand based on my talk in NYC built around the concept of morphogenesis, hence Morphxgen. That presentation helped solidify my focus and positioning within the ever-growing world of Computational Design.
Morphxgen formalizes – what were once my side projects – into a creative outlet to put real objects out into the world.
In Tokyo I will be sharing a brief glimpse into the early stages of this brand, as well as some possibilities for the future of additive footwear. I see a lot of opportunities in this space that not many brands are taking advantage of, so I thought I would throw my hat into the ring so to speak. I will also be showing a few of the features of my footwear creation environment that I have been building for some time now.
In New York the focus was on temporal design and simulation as a design tool, and in Barcelona it was a production toolchain for a non-technical operator. How do those two threads come together in the footwear work you will be presenting?
In short, the work I am presenting in Tokyo combines many concepts from both these former presentations. I am still ingesting and analyzing temporal data to drive computational systems, I am leveraging custom solvers and simulations for geometry creation, and I am again tackling the all-in-one design and production eco-system, but this time the user is, myself.
I am continually interested in how the human body works and how footwear is an interface with which we interact with our world, so I will continue the thread of biomechanics and locomotion as well.
The abstract asks how custom additive footwear really is. What prompted that question, and what does customization mean in practice?
Scan-to-fit, bespoke, 1-of-1, these are phrases that get thrown around in the footwear space quite a bit. It is something that probably every major brand has tackled or at least considered at some point. But is that the limit of customization?
I think this is actually only the beginning, the most basic form of custom which is not that much different from how footwear was made pre-industrialization, by taking measurements of your feet.
I believe there is a lot more meat on the bone in terms of how footwear can be made for an individual human or athlete. Beyond just the 3D geometry of someone’s foot there is gait cycle analysis, how the foot bends, expands, contracts, how much force someone applies to the ground, how asymmetrical they are, their individual locomotion patterns. These are all ways in which motion and data can be translated into custom geometry.
Houdini has been the backbone of your previous work. Is that still the case? How does data move from a body or performance input through to a printable structure?
Yes, nearly 100% of my process for Morphxgen is built in Houdini. From automated size grading, procedural lattice generation, geometry analysis, temporal data processing, data-driven design, generative algorithms, simulations, visualization, etc. I need a singular environment that can be configured to do it all, and as of today Houdini is the most capable software for the job.
The ways in which data can be translated into printable geometry is where I suppose the crux of this work lies.
I often consider the computational design work I do to basically be summed up as translating human biomechanics or performance goals into geometry. For this project specifically I am looking at how simulations from gait cycles can drive stress-tensor extraction to feed anisotropic lattice structures.
Basically I want to see if motion can create printed footwear that moves the way our bodies move, or even contributes to better movement.
Multi-agent algorithms and stochastic lattices can produce a wide range of outputs. How do you keep that variability within manufacturing constraints, and does “mostly good, basically all the time” still hold as the standard?
Part of my choice for pursuing multi-agent systems and anisotropy was actually because there is a fairly small range dimensionally within elements of a lattice structure where both manufacturability and mechanical properties are in balance. Particularly with SLA and light-based resin systems, beam or wall thickness might only be able to vary by say about 1mm.
Demos where we see extreme gradients in print thickness likely would not work in this environment because slightly too thick and your material becomes extremely stiff, slightly too thin and you will have problems printing these features.
I needed other ways in which I could control the mechanical properties of complex lattice structures, so I targeted anisotropy, beam direction, and connection architecture through multiagent and other systems such as our old friend voronoi. It did take many rounds of prototyping to reach a state where I think the footwear is providing something that no other company is doing at the moment, at least in the additive space. The flexibility and energy absorption in particular are unlike other printed footwear.
What do you hope people take away from the presentation, and what are you hoping to learn from the other presenters and attendees in Tokyo?
I hope people expand their horizons of what is possible within the space of additive manufacturing for footwear.
I want to challenge the perception of how additive footwear can feel, and look.
I am hoping to learn about others who are building bespoke products at scale, other automated methods of make, and check out the general design and engineering scene in Japan.
Join us in Tokyo at CDFAM to hear directly from David and other leading experts in computational design at all scales, from architected materials to architectural systems, October 8-9, 2026 at TOKYO INTERNATIONAL FORUM.
