Running Revolution: Computational Design Behind PUMA Fast-R NITRO Elite 3
Interview with Jesus Marini Parissi of Moon Rabbit Lab X PUMA
Could you start by introducing Moon Rabbit Lab and describing what you will be presenting at CDFAM about the Fast-R NITRO Elite 3 project with PUMA?
Moon Rabbit Lab is a computational design consultancy based in Milan, Italy. Our core expertise lies in performance-driven design, with a focus on sports footwear, outdoor sports, and advanced consumer goods.
We’ve developed a workflow that enables companies to translate real data into measurable product improvements, whether that means faster development cycles, reduced material waste, or, as in the case of the FAST-R 3, a significant boost in performance.
At NYC, we’ll showcase how computational Design played a key role in elevating the FAST-R 3, driving a 3% improvement in running economy.
By combining insights from previous models with advanced virtual simulations, we transformed data into a leap forward for performance footwear.
How does your computational design workflow integrate digital simulation, biomechanical analysis, and optimization, and what were the key factors in aligning these elements effectively?
That’s a great question. Let me give you a bit of context.
When working on innovation projects, it’s crucial to think systematically: understanding the broader context, identifying connections between different domains, and building the right “bridges” to make them communicate effectively.
In our case, the challenge was to converge digital, biomechanical data, and computational optimization. To do so, we needed an environment capable of handling all these dimensions simultaneously. That’s why, from the very beginning, we worked in a 3D design space rather than the traditional 2D approach. This choice allowed us to map every type of data directly into the design environment, helping us identify optimized volumes and configurations.
In practice, we assign properties to each element within the design space, performance targets, material behaviors, and manufacturing constraints, and arrange them to achieve the best balance.
The critical factor here is not only computation, but also keeping humans in the loop: using digital results as guidance while remaining critical, selective, and driven by real-world performance goals.
What software tools and simulation platforms were used to run and evaluate the hundreds of virtual iterations, and how did they interact within the design process?
I’d like to approach this question from a different angle. At Moon Rabbit, we see ourselves first and foremost as problem solvers.
Our mission is not to force a challenge into an existing tool or software, but to deeply understand the core of the problem and then identify, or create, the most effective way to solve it.
Often, the right solution doesn’t exist in the market, so instead of adapting the problem to the tool, we build the tools that fit the problem.
This allows us to design our workflow as a system from the start, where every part of the process interacts seamlessly. We break it down into modules, making it easier to test, and optimize quickly.
Sometimes this means investing significant effort upfront, but that investment pays off by saving time, reducing iterations, and delivering more reliable outcomes at the end.
Can you outline the data flow from initial performance KPIs through simulation results to final design decisions, and how you validated these before physical prototyping?
From the start, our KPI was clear: reduce weight without compromising performance, while preserving the identity of the FAST-R line. To achieve this, we first analyzed the previous iterations, FAST-R and FAST-R2, extracting insights and learnings from each. In parallel, we studied the running gait in detail.
Even though much of the work happened in a virtual environment, we didn’t abandon traditional steps: sketching over technical drawings, refining curvatures, proportions, and volumes, and carefully matching materials and hardness levels to the right zones. This critical, iterative thinking ensured the design was both technically sound and performance-oriented.
Collaboration with the PUMA Innovation team was essential to define an aesthetic language aligned with PUMA’s DNA. On the technical side, we built an efficient virtual environment able to integrate biomechanical data while ensuring materials behaved realistically in simulation.
Together with PUMA, we invested significant effort in accurate material characterization and digital modeling so that our simulations would closely mirror reality. Finally, we validated the FAST-R 3 digital model by correlating it against physical data from FAST-R2, ensuring continuity, credibility, and measurable progress.
What challenges did you face in balancing weight reduction with performance gains, and how did computational optimization guide trade-off decisions?
It was a complex task, as we had to consider multiple aspects simultaneously: manufacturing, biomechanics, aesthetics, performance, foam decap, and carbon plate stiffness.
To achieve the right balance, we went through several virtual design iterations, validating some ideas and discarding others. Digital simulations and energetic analyses provided the insights needed to make informed decisions. Digitally, we were able to test different material configurations, identifying those that delivered optimal performance and the most efficient energy flow.
Computational design played a crucial role in shaping the carbon plate. Using the tools we developed, we were able to precisely calibrate spacing, orientation, and rib design down to the millimeter.
The most rewarding moment for me was completing the carbon plate design.
Seeing the 3D model, checking all tangent lines, edges, and G1 continuity, it all looked perfect.
For me, it was a true harmony between form and function, and that was the highlight of the entire process.
What do you hope to share with, and learn from, other participants at CDFAM about applying data-driven computational design across performance-focused consumer products?
I’m looking forward to learning how other industries are applying computational design and understanding the key challenges of introducing these methodologies into the development process.
It’s also a great opportunity to expand our network and share Moonrabbit’s vision and mission to inspire both new and established generations to take part in this new era of design and engineering.
To learn more about Moon Rabbit’s design process, and connect with other design and engineering experts from footwear and beyond, register to attend CDFAM in NYC October 29-30, 2025 for two days of networking and knowledge sharing with leading experts in computational design and AI/ML for engineering at all scales.
