0:00 So one of the tools that I built was something that can kind of take in all of this data this motion data and actually visualize and analyze some of it over time but can also extract some really interesting kind of approximations. So looking at someone’s center of mass looking at someone’s symmetry and also looking at someone’s ground reaction force which is basically how much force they’re imparting into the ground when they run, walk, jog, etc.

0:30 Recorded at the CDFM computational design symposium in Tokyo, October 2026. Learn more at CDFAM.com. My name is David Burpie. I am a computational designer. And I’ve been working in this space for a little over 10 years. I currently work for a company called Brooks Running. We make running shoes. I work on the elite athlete innovation team. So working on kind of the next suite of running shoes of the future.

1:07 I also have my own design practice for the past few years. And I’m launching a small project called Morphagen, which I’m here to talk a little bit about today. I’ve also been an educator and a researcher for a number of years. Have lectured and taught at a number of universities including the University of Washington, in Seattle where I’m based out of. And I’m kind of a hobby algorithmic art and kind of design nerd.

1:31 Super interested in emergent design systems, biographic simulation and things like that. And these are a few of the brands that I’ve been lucky enough to work with over the years. So I want to talk a little bit about my background. I think you’ll see some threads through my work and how they kind of get expressed in the project. So originally I was a computational designer in the world of architecture.

1:59 I worked primarily in the high-rise and super tall kind of realm. So I have a number of projects in Asia as well as kind of up the western seabboard in the United States. And as I mentioned, I’m also super interested in algorithmic art kind of generative systems. So began many years ago in a software kind of IDE and language called Processing which is built on Java, but I’ve also used Touch Designer, Houdini and kind of a number of other tools as well.

2:27 And this always was kind of a creative way for me to to get into coding and just kind of a more expressive way to do what I consider the fun parts of of computational design. And as I mentioned earlier, I’ve been a part of a number of kind of research initiatives and academic initiatives. This was one in particular that I thought was pretty interesting that I talked a little bit about at my first CDFM in New York City.

2:53 And it’s around kind of designing and using engineered living materials which is embedding living organisms, bacteria, things like that into 3D printable polymers to elicit certain behaviors like self-strengthening, self-hen things like that. This was a multid-disiplinary multi-university university project that was funded by the NSF. And for the better part of I would say five years I’ve been working mostly in the world of footwear. So have worked for a number of brands as I mentioned earlier.

3:27 Exploring a number of different kind of computational design flavors or technologies in the world of footwear. Specifically I’m super interested in this world of kind of algorithmic and biomimetic design. So I have a number of projects over the years that have kind of incorporated these things into my own work. Some of which are seen here and some of which I’ve talked about before. Today I’m prim primarily going to be talking about the specimen zero project which is on the left.

3:58 So show of hands, who here has like side projects that they work on in their free time. Yeah, most people right. I think we all have probably too many side projects, especially the people in this room. I myself had too many as well. And I kind of got to a point where I felt like I needed to actually put some of this stuff out to the world.

4:17 I think you can only get so far kind of noodling and exploring on your own at home. So I decided to start actually a kind of small brand. And the brand is called Morphagen. And the brand started actually kind of out of my presentation in New York City where I looked at this concept of morphagenesis. If you’re familiar or if you’re not familiar with it, it’s a concept that’s coined by Alan Turing and it’s really how structures and things in nature get their shape and get their form.

4:51 So, like I’m sure some of you, I’ve always been fascinated by kind of natural systems, how they can self-organize and self-optimized and also I think just some of the beautiful structures and patterns that are seen in nature. And have had the kind of benefit of millions of years of evolution to to get to where they are. So it’s always been kind of an endless muse for for me and my work.

5:16 So today I’m going to talk a little bit about many of you probably know the adage by Louisis Silivven form follows function. What I’m actually interested in is how form can follow force. So, if you’re familiar with the kind of world of footwware design, particularly additive footwear, you may have heard this term called scan to fit, which is essentially taking a 3D scan of someone’s foot, and then making a kind of quote unquote bespoke footwear for them.

5:44 It’s also seen in a couple of other other industries as well. To me, that is kind of asking the question like what shape is someone’s foot? And what I’m actually interested in is how does this person move? How do they kneel, squat, run? How do they get tired? What kind of forces do they put into the ground? To me, this is a much more kind of compelling and interesting use case to start to design additive structures around.

6:07 And I’m referring to this as kind of a biomechanical footprint or fingerprint rather. So, the overall architecture of the system I’m going to walk through today is as follows. There’s a human data capture element which I’ll show you in a second. From this, I’m going to generate what’s called a stress tensor field. I’m going to use some computational techniques to then process this tensor field into a simulationdriven lattice geometry, which is then print ready to go to any kind of major 3D printing manufacturer.

6:42 So given that this is just kind of a side project of mine, I was looking for some kind of open-source sources to see if I could start to look at some biomechanics data. Carnegie Melon University happens to have a pretty large catalog of what’s called motion capture data. You’ve probably seen this used in like film or gaming. The premise is you pretty much put these little markers or identifiers on someone’s body.

7:08 You then have a room with cameras or sensors that then can track their movement over time. And Carnegie Melon has a series of different kind of movement scenarios like running, walking, jogging, etc. So this data set became kind of my exploration and test data set. And I’m also using a little bit of my own data as well. So, if you take kind of the raw data, which is really just numerical, it’s just the location of all those markers, and you start to rig a skeleton and you kind of animate this, and you can look at how this person is actually moving in space.

7:45 And this creates kind of a templated version of this person’s body. And it just shows kind of their progression of motion through time. And I was considering, okay, like how can I use this and what can I get out of this? So one of the tools that I built was something that can kind of take in all of this data this motion data and actually visualize and analyze some of it over time u but can also extract some really interesting kind of approximations.

8:13 So looking at someone’s center of mass looking at someone’s symmetry and also looking at someone’s ground reaction force which is basically how much force they’re imparting into the ground when they run, walk, jog, etc. And at first it was kind of a visualization tool, but it slowly progressed into something a little bit more sophisticated where I was actually able to get kind of numerical data out of it.

8:35 Obviously, it also works with other kind of movement styles. This is someone boxing, I think. But I thought it was kind of interesting as a process to start to leverage a kind of ecosystem that’s built around human data and motion capture. Another really interesting thing about these skeletons is that they are used in the world of film and VFX and you can actually then rig certain geometry onto them.

9:05 So in this case at the top I actually have a scan of my own foot which I then rigged onto one of these skeletons. In an ideal scenario you would probably have your own 3D body scan that’s correlated to your own kind of motion capture data but I didn’t have that. And what I’m doing here is I’m essentially just showing a single kind of walk cycle or gate cycle of this geometry.

9:30 And I want to understand how is this actually deforming, how what areas of the foot are expanding, what areas are contracting. So the kind of mesh deformation that you see on the screen with the kind of false color so shows certain areas during the walk cycle that are experiencing more or less of this kind of expansion and contraction. From there I can measure the gradient of these changes and I can actually extract a tensor field on the surface of this geometry to understand the kind of directionality of these forces.

10:04 Another common source of kind of temporal data or human data in footware design is something called a pressure map. You may have seen these before. They can be static, they can be dynamic, they can be captured a few different ways, but it’s another way to kind of understand how someone’s body is interacting with the ground or interacting with kind of a shoe interface. So, this was another kind of data set that I was leveraging for some of this work.

And I wanted to kind of create a type of footwear that is maybe a little bit different than what people are are used to in the world of additive footwear. This is kind of a a tech booty. I would say it’s there’s a term in footwear called recovery shoe, which I think is a bit of a nebulous term, but it probably fits into this category. It’s something that you can kind of wear every day.

10:52 You can move in, you can walk your dogs, take your kids to the park. But it’s kind of a generic piece of footwear. And I’m calling it specimen zero because it is really kind of the baseline. And it was really just a platform for me to begin to test some of these these problems and kind of the usage of this data with the eventual goal being to build up to other kind of more sophisticated types of footware.

11:16 So the thing I was looking to explore was to see if I could use this human data to actually drive anotropy in lattice structures. So again, I’m sure many people in this room are familiar with anisotropy, but in a very simple sense, it’s just mechanical properties that kind of deviate based upon direction inside of a structure. So wood grain is a really simple example. We all know that it’s easier to split wood, you know, with the grain rather than against the grain.

11:42 And we also have these kind of structures actually inside of our own bones. So these are called tbecular structures that you see on the right. And what I think is really interesting is you can almost trace the load path kind of through this femur and you can see that these structures have been kind of laid out in a way that is already expressing the kind of loading case of this bone.

12:06 So I was looking for different types of lattice geometries. I settled on our friend Boroi for a number of reasons. It’s a pretty common stochastic lattice structure that I’m sure many of us have used before. And but what I wanted to do was kind of push it a little bit further and do it in such a way where I could actually create these cellular structures that are growing and changing based upon this kind of underlaying underlying vector field.

12:34 So these were some early studies of of prototypes that I built. These as well as all the kind of lattice creation I was building in Houdini. So this is kind of a simulation that basically ingests this kind of scalar field and based upon the density, the magnitude and the direction of these vectors will kind of settle on an anisotropic structure that uses kind of an underlying modified voronoi algorithm.

13:05 So this is that same kind of process running on the actual kind of footware geometry. You can see some of the features that I have highlighted of creating kind of like this almost strap idea almost like a sandal and kind of leveraging this anisotropy to create kind of pseudo features within the footwear itself. And then again you can see the kind of underlying pressure map data on the bottom as well and how that’s driving the structure.

13:31 And then the process was actually built procedurally. So if this video will play but I’m basically using that kind of stress tensor field on the surface of this geometry building up layers of this borono pro diagram and then creating kind of an intermediate lattice structure between the two that’s all driven algorithmically and procedurally. This doesn’t want to play for some reason. One of the other kind of challenges in footwear particularly additive footwear is this idea of size grading.

13:58 So traditional shoes generally get designed for one size and then are kind of graded up and down the size spectrum. What’s different about additive footwear is you usually have to compute a new lattice structure for every size or every other size. You can’t just simply take a single lattice structure and kind of scale it up and down to meet all the different size needs from like a woman’s four to like a men’s 12.

14:22 So this was all built procedurally. So it was kind of a parameterized version of this base form. That I can just pick what sizes I want to kind of execute at any given time and it automatically computes the size grading for these and then I’m procedurally generating the lattice structure on top of it. And I’ll talk a little bit about validation as well. The general idea is that I would like to feed this kind of lattice creation back into the kind of biomechanics portion.

14:52 For those of you that may work in the footware space, you probably know that simulating a full human body is actually a very complicated task. So I started with doing just kind of simple compression and deformation and bending analysis. The idea is that at some point it would be a little bit more closed loop and it would start to feed back into the beginning of the cycle.

15:15 That whole system is a little bit still in progress, maybe not ready for prime time yet, but perhaps a future CDF fan presentation. But one of the best ways to actually kind of understand these structures, see if they work, see if they feel interesting is to actually just physically prototype them. And this is the one of one of the benefits of using additive manufacturing is they’re fairly quick to send a model over and get a 3D printed structure back.

15:41 So I’ve gone through a few prototyping iterations already. I’ve obviously worn them many times. And kind of understand I think once you hold them in your hand you can intuitit a lot of whether or not these behaviors are actually occurring. So these are a few images of one kind of prototyping round. I think what’s interesting is you can kind of start to see the flow. The kind of diagram of forces is almost visible in the final structure and particularly when you look at kind of the outsole view you can almost kind of reminisce and see like oh the pressure map is actually kind of creating a variety of different structures in the footwear which I personally think is really interesting and then here’s a few different colorways as well that I was working with and I also have some there’s some more complexity to kind of the algorithm building of the lattice geometry.

16:33 So, there’s certain zones like footed areas and different kind of dimensions that I’m working with. But ultimately it’s all kind of around this goal of how you process human movement data and how you translate that into geometry. And then a photo of some packaging there. I actually have some with me today. So, if you do want to come check them out, just find me afterwards. And if you happen to be a US men’s nine, you can even try some on and walk around in them.

17:06 And so the ultimate goal is really to have this kind of closed loop human data capture system. When I started this project, it was really revolving around how do I drive lattice creation. The more I got into it, the more I became interested in how do you actually use what is essentially very complex and temporal data of how a human moves, runs, etc., and can create a closed loop system that can make optimized lattice structures out of this data.

17:36 If you want to get in touch with me, here’s some socials. Otherwise, you can come find me afterwards. And thank you very much for your time. Recorded at the CDFM computational design symposium in Tokyo, October 2026. Learn more at CDFAM.com.