Footwear has always been a deeply physical discipline. You stretch leather, feel foam, bend plates, hand-shape uppers over lasts. So what does it mean when design becomes computational, when the raw material isn’t leather or EVA but logic, data, and systems thinking?

For the past decade, “digital transformation” in footwear has meant CAD, rendering, and shrinking the sample cycle. But computational design goes further. It reframes the entire problem. It shifts the designer’s role from shaping forms to defining rules, behaviours, and relationships. It moves footwear from object to ecosystem.

Not asking “how do we draw this better?” but “what rules create it?”
Not asking “what does it look like?” but “how does it behave?”

Recently, we brought together some of the most advanced computational thinkers in footwear; people who are blurring the lines between engineering, craft, robotics, simulation, and design:

Onur Gün, Director of Computational Design, New Balance
Jesus Marini Parissi, Founder, Moon Rabbit Lab
Maia Zheliazkova, Senior Computational Designer, On
Eric Hullegie, Head of Innovation Concept Design, On
René Medel, Senior Digital Creation Engineer, framas Group

🎥 Watch the full discussion below, and read on for the ideas reshaping how footwear gets imagined, modelled, tested and made.

Key Themes

1. Computational design is a mindset

Footwear has historically been “material-first.” You determine the substance, then shape it. But computation reverses the flow: instead of starting with geometry, it starts with relationships, behaviours, constraints, and rules.

Computation acts as connective tissue, linking performance intent, material engineering, automation, and the final expression.
- Maia Zheliazkova, On

It’s less about geometry, more about ecosystems.

What architecture learned a decade ago, moving from drawing buildings to modelling systems, is now happening in footwear. Designers aren’t just asking “how do we make this shape?” but “what generates this shape, and why?”

Bottom Line: The power of computation lies not in scripts, but in the new cognitive model it enables.

2. Adoption is slow

Footwear design is inseparable from footwear manufacturing, and manufacturing has evolved slowly. Manufacturing constraints, mould costs, material variability and long-standing workflows anchor an industry to the familiar.

How we design is extremely tied to the manufacturing processes we use, and those have been static for a long time.
- Eric Hullegie, On

Computational design only accelerates when manufacturing evolves in parallel through robotic deposition, adaptive moulds, 3D printing, automated upper formation and advanced material characterisation.

The other barrier? Talent. Computational footwear designers need hybrid fluency - code, geometry, biomechanics, materials, systems - but few educational pipelines exist.

There are not many programmes for this. You need design, engineering, coding, systems…all in one.
- Jesus Marini Parissi, Founder, Moon Rabbit Lab

Bottom Line: Adoption isn’t blocked by the tools, but by the organisational and industrial systems wrapped around them.

3. Data is powerful, but incomplete

To design computationally, you translate reality into data: gait cycles, pressure maps, fatigue curves, material responses and more. But data alone can’t capture material nuance, manufacturing quirks, or the lived knowledge of how shoes actually come together.

Data can reveal patterns, but it cannot replace hands-on manufacturing knowledge.

If you don’t know how to make shoes, the data won’t save you.
- René Medel, framas Group

And Jesus emphasised the hidden labour of data: labelling, cleaning, filtering, checking accuracy, understanding what’s missing.

Data collection is the hardest part - and if the quality is not good, you need to shrink the solution space.
- Jesus Marini, Moon Rabbit Lab

Eric added that before data, footwear development was almost entirely empirical: prototypes, gut feel, athlete feedback. Today we can simulate much more, but reality is still messier than the models.

Bottom Line: Data guides decisions, but intuition interprets the gaps.

4. Simulation vs. sensation

The holy grail: predictive models that simulate wear, deformation, comfort, responsiveness and long-term performance. We're getting closer, but we’re not there yet.

Simulations rely on simplification. Every model requires reducing the real world into a smaller, solvable subset: material behaviours, environmental conditions, biomechanics, variability and time. Accuracy increases, but human sensation still defies perfect capture.

We try to build this iterative loop between simulation, material behaviour, prototypes, and human feedback - but each speaks a slightly different language.
- Maia Zheliazkova, On

This friction between the measurable and the felt may always exist.

Bottom Line: Simulation sharpens human insight, but won’t replace it.

5. Craft resurges through automation

Industry-wide, automation is seen as a threat to craft. But for computational designers, it’s the opposite.

Automation removes repetitive tasks and allows designers to focus on the complex: behaviours, interactions, system boundaries, and new manufacturing possibilities.

Automation gives us time. It doesn’t replace us — it lets us design the next workflow.
- Jesus Marini, Moon Rabbit Lab

And the new frontier is mass customisation: adaptive moulds, programmable forms, 3D printed components, robotic deposition systems. Each reduces the gap between design intent and physical output, but only if we protect the tacit knowledge of how shoes are really made.

The times of forecasting hundreds of millions of pairs may be ending - production is decentralising.
- René Medel, framas Group

Bottom Line: Automation expands craft by shifting human attention toward the decisions that matter.

6. AI accelerates creativity

Most people talk about generative AI - the image-making, concept-sketching, idea-spitting surface layer. But in computational design, AI is a collaborator of a different kind.

AI is becoming:
- a compiler for designers who don’t code
- a co-pilot for algorithm development
- a pattern recogniser across materials and biomechanics
- an accelerator for testing ideas quickly and cheaply

I have no coding background, but now I have the power of an expert coder in my hands.
- Eric Hullegie, On

But AI is not an autonomous creator. Its limitations - hallucinations, gaps in training data, lack of physical grounding - matter deeply in footwear, and cannot replace grounding in real materials and manufacturing.

AI can’t handle newness yet...that’s still us making the call.
- René Medel, framas Group

Bottom Line: AI makes designers faster, but it can only amplify the quality of the questions we ask.

7. Cross-disciplinary by design

The best computational designers are rarely purists. They are generalist specialists: architects, engineers, coders, shoemakers, roboticists, biomechanists.

Architecture teaches system thinking.
Engineering teaches constraints.
Gaming teaches real-time simulation.
Footwear teaches material reality.

When these collide, breakthroughs emerge, like On’s LightSpray, New Balance’s data-driven midsole work, or framas Group’s advancements in automated mould creation.

Bringing architecture and engineering mindsets into footwear is where breakthroughs happen.
- Onur Gün, New Balance

Bottom Line: The future belongs to teams who mix disciplines, not protect them.

Computational design won’t replace designers any more than simulation will replace athletes. But it will compress idea-to-prototype cycles, align manufacturing and design like never before, enable new forms and behaviours that can’t be sketched, expand what materials and machines can express, and reveal where automation ends and human judgment begins.

At its best, computational design turns sensation into structure, data into decisions, and constraints into creativity.
- Onur Gün, New Balance

We’ll continue exploring this space at Stride USA (Portland, March 2026) and Stride Europe (Italy, May 2026) - bringing together designers, engineers, technologists, and makers shaping the next era of product creation. Join us.

Stride USA 2026: https://apparel.pi.tv/events/204/agenda
Stride Europe 2026: https://apparel.pi.tv/events/208/agenda