Barcelona
April 8-9, 2026
The Barcelona program will feature two days of technical presentations and case studies spanning computational design, simulation-driven engineering, advanced manufacturing, and architectural systems.
The agenda highlights the emerging role of AI and machine-learning–based methods within these practices, with contributions from industry, academia, and software developers focused on practical workflows, underlying computational frameworks, and real-world application.
Supporting Sponsors
Program Archive
Organization:
CDFAM
Presenter:
Duann Scott
Welcome to CDFAM Barcelona
Presentation Abstract
Welcome to CDFAM Barcelona and opening remarks giving context to the event and it’s program.
Speaker Bio
Duann Scott is the founder of the CDFAM – Computational Design Symposium, bringing together experts across design, engineering, and software development to explore the future of computational design at all scales.
He is also the Executive Director of the 3MF Consortium, driving the development of an open data standard for additive manufacturing.
With a background spanning industry, academia, and software, Duann focuses on advancing innovation in digital design and advanced manufacturing.
Organization:
NVIDIA
Presenter:
Pablo Hermoso Moreno
Accelerating CAE: AI Physics, Surrogate Models, and Agentic Workflows
Presentation Abstract
Computer-Aided Engineering (CAE) has fundamentally transformed product development, yet designing complex systems remains bottlenecked by time-consuming, highly manual workflows and resource-heavy computations. This presentation explores how NVIDIA addresses these challenges by fusing accelerated computing with Artificial Intelligence (AI) Physics to drastically cut down simulation solve times. By deploying AI surrogate models, we provide engineers with high-fidelity, real-time feedback, bridging the gap toward fully responsive digital twins.
To tackle the manual bottlenecks, we will introduce the transformative capabilities of Agentic AI Engineering powered by the NVIDIA NeMo Stack. We will explore practical examples of autonomous agents automating end-to-end CAE workflows—slashing the time required for requirements gathering, simulation setup, and post-processing analysis.
For instance, we will look at the AI-Q Blueprint, a production-ready enterprise research agent built using LangChain Deep Agents. We will detail how the NeMo Agent Toolkit (NAT) profiles agent behavior to tag each Large Language Model (LLM) call with latency sensitivity hints, and how Dynamo leverages these hints to let high-priority requests jump the queue, drastically reducing interactive latency without requiring new hardware.
Speaker Bio
Pablo is a Senior Solutions Architect at NVIDIA focused on AI for physical simulation, surrogate modelling, design optimization, and agentic systems for CAE workflows. He holds a degree in Aerospace Engineering from Imperial College London and conducted research at Caltech on Physics-Informed Machine Learning. Previously, he worked as an Aerodynamics Engineer at Mercedes F1 before joining AWS as a Senior Data Scientist, where he delivered scalable AI/ML solutions for large enterprises on the cloud.
Artificial Intuition: Building an AI Mind for Electromagnetic Design and Engineering
Presentation Abstract
Arena is building an AI foundation model for electromagnetism to accelerate our rate of innovation in electronic hardware – from radios to radars to computers. Arena works with the leading companies across the semiconductor, aerospace and automotive sectors. Arena is backed by $62 MM in Venture Capital from investors including Peter Thiel, Garry Tan, Founders Fund, Initialized Capital, Goldcrest capital and notable others.
Speaker Bio
Mike Frei earned his PhD in physics from Columbia studying how geometry governs electron behavior at the molecular scale, then spent a decade at McKinsey and in C-suite roles at RUAG, a Swiss aerospace and defense company, developing a rare combination of deep physics intuition and operational experience in the systems Arena Physica is now reimagining. Michael brings 10+ years of experience driving large scale business transformations and operational programs in multi-billion dollar international companies, with experience serving ~15 different clients at McKinsey. More recently Michael worked on Board and ExCo level for a high-tech Space and Defense player, overseeing the re-design of entire operational units. Michael holds a Ph.D. and M.S. in Physics from Columbia University and a B.A. in Physics from University of California, Berkeley.
Functional AI for 3D Design Automation — From Path Finding to Generative Modeling for Building Construction
Presentation Abstract
Great strides made recently in 3D generative artificial intelligence (GenAI) have been propelled by the rapid scaling of large foundation models and advances in generative models such as diffusion and flow matching. However, current neural generators have predominantly been constructed by optimization against image-space losses. Should appearance be the main criterion for 3D design and content generation? Not really. The 3D world we live in is not only to be observed. Accordingly, the main goal for 3D GenAI should be for the generated 3D entities to be used and interacted with, so as to serve their intended functions, just as in the real world.
We introduce Functional AI to 3D design automation and demonstrate its importance and potential for the built environment. Fundamentally, any constructed building, and all the objects and structures therein, must fulfill the desired functional requirements, from architecture and complex structural layouts down to the placements and intricate interplay between mechanical equipment, heat or water pipes, and electrical conduits. Our technical coverages will encompass functionalization of 3D objects and scenes, agentic AI for path finding, and generative modeling of complex building structures, with the ultimate goal of establishing a foundation model for building data with construction intelligence.
Acquiring the requisite semantic and spatial understanding, and developing GenAI tools to attain functional designs, is especially challenging for non-residential constructions, specifically commercial, medical, institutional, and mission critical buildings such as data centres. This is the hard problem Augmenta has been focusing on. Our presentation will showcase recent successes of bringing large-scale design automation and generative design to practice, highlighting two elementary schools in Michigan – the first buildings in the world with an electrical system modeled and delivered by AI-powered generative design.
Speaker Bio
Hao (Richard) Zhang is Vice President of AI and R&D at Augmenta, and a Professor in the School of Computing Science at Simon Fraser University, Canada. Augmenta is a Canadian start-up that aspires to automate building designs, from mechanical and electrical to structural and architectural. Richard is leading the company’s efforts in developing advanced AI tools to achieve these goals with scaling, efficiency, and sustainability
Richard is a Fellow of the IEEE, holds a Distinguished University Professorship, and is a member of the ACM SIGGRAPH Academy. From 2021 to 2025, he was an Amazon Scholar. Richard obtained his Ph.D. from the University of Toronto, and MMath and BMath degrees from the University of Waterloo. His research is in computer graphics and visual computing with special interests in spatial and functional AI, geometric and generative modeling, 3D vision, geometric deep learning, as well as computational design and fabrication. Awards won by Richard include a Canadian Human-Computer Communications Society Achievement Award in Computer Graphics (2022), a Google Faculty Award (2019), an NSERC Discovery Accelerator Supplement Award (2014), a Best Dataset Award from ChinaGraph (2020), as well as faculty grants/gifts from Adobe, Autodesk, and Google. He and his students and collaborators have won the SIGGRAPH 2025 Test of Time Award, CVPR 2020 Best Student Paper Award, and Best Paper Awards at Symposium on Geometry Processing (SGP) 2008, and CAD/Graphics 2017. Richard was the Technical Papers Chair for SIGGRAPH 2025.
Organization:
Universidad Europea de Madrid
Airbus Operations SL
Presenter:
Raul C. Llamas-Sandin
Unlocking Large-Scale Structural Synthesis: High-Performance GPU Topology Optimization for Architectural and Civil Engineering Applications
Presentation Abstract
This presentation introduces a novel, high-performance topology optimization solver developed to bridge the gap between computational efficiency and the massive scale required for architectural and civil engineering projects. While traditional topology optimization is often limited by mesh resolution and computational cost we leverage the massive parallelism of GPUs to solve structural synthesis problems with millions of elements in minutes on a workstation.
Key features relevant to the built environment include robust stress constraints with asymmetric tension-compression limits, active thermal loading for thermoelastic analysis, and the ability to enforce strict geometric constraints using external geometry to define non-designable void or solid regions—essential for integrating fixed features or architectural fenestration.
For civil engineering, we showcase applications in bridge design and high-rise structural cores, where the solver optimizes material distribution within large domains (e.g., 80m spans) under complex loading scenarios. In architectural design, the software enables the synthesis of free-form, structurally efficient forms that are immediately exportable as watertight, smoothed STL files for digital manufacturing or rapid prototyping. By decoupling the physical problem definition from the numerical solver via modular configurations, the software offers a flexible, production-ready workflow that empowers designers to explore the limits of structural efficiency without compromising on resolution or speed. This tool represents a significant step forward in democratizing high-fidelity structural synthesis for the Architectural and Engineering design.
Speaker Bio
Raul C. Llamas-Sandin is an aerospace engineer and academic with over two decades of experience in advanced aircraft design and technological innovation. Since 2001, he has served as a Future Projects Engineer at Airbus, where he specializes in conceptual design and leads research initiatives in European R&T framework programmes.
Parallel to his industrial career, Raul has been an Assistant Professor of Aerospace Engineering at Universidad Europea since 2011, teaching courses on Flight Mechanics and Advanced Aircraft Design. His technical expertise spans aerodynamics, structural engineering, and preliminary sizing—roots established during his early tenure as a Structures Engineer at BAE Systems in the UK.
Raul holds an MSc in Aerospace Vehicle Design from Cranfield University, an MSc in Physics from UNED, an Aerospace Engineering degree from the Polytechnic University of Madrid and is currently pursuing a PhD in Mechanical Engineering at the University of Seville. He holds 14 patents in areas such as aerodynamic improvement devices and plasma actuation, and authoring numerous publications on topics ranging from AI in structural sizing to experimental ice shape analysis.
Organization:
Fraunhofer Institute for Computer Graphics Research IGD
Presenter:
SubSimX: Interactive Subdivision-to-FEM for Computational Design
Presentation Abstract
Computational designers increasingly work with lightweight, organic geometries, and vast design spaces, but the connection between modeling and simulation is still constrained by slow, fragile meshing workflows. Every shape change requires rebuilding a volumetric mesh and redefining boundary conditions, which breaks interactive exploration and limits the practical use of shape and topology optimization, generative methods, and AI-based design in daily engineering practice.
We present SubSimX, a subdivision-native design and simulation pipeline that turns Catmull-Clark subdivision models into a live, structurally aware modeling environment. Designers work directly on the familiar subdivision control mesh, while volumetric meshing and boundary-condition setup are performed only once at the beginning of the process. Subsequent edits to the control mesh are transferred to the existing analysis mesh within seconds, without invalidating boundary conditions, so geometry, performance, and manufacturability can be explored together in a continuous, simulation-driven workflow.
At the core of SubSimX is a precomputed mapping that describes how the interior of the part follows movements of the control mesh. This mapping enables fast updates of the tetrahedral analysis mesh whenever the geometry changes. A lightweight, local mesh-repair step maintains element quality during large deformations while preserving mesh topology. Coupled with RISTRA, our fully GPU-accelerated finite element solver, this delivers real-time or near real-time structural feedback directly inside the design session.
For the computational design community, this effectively turns subdivision modeling into a responsive structural design medium. Dense manual or scripted design sweeps, topology optimization followed by smooth subdivision refinement and shape optimization, or AI-in-the-loop exploration become feasible within industrial workflows, including casting-oriented lightweight structures. We demonstrate how SubSimX reduces iteration times from hours to seconds and closes the loop between exploratory design and real-world engineering.
Speaker Bio
Johannes Mueller-Roemer is deputy head of the Interactive Engineering Technologies department at the Fraunhofer Institute for Computer Graphics Research IGD, a role he has held since 2023. His work focuses on GPU-accelerated algorithms and data structures that tightly couple geometric modeling with interactive simulation, enabling computational design workflows for generative, optimization-driven engineering.
He has been a research scientist at Fraunhofer IGD since 2011 and received his PhD from TU Darmstadt in 2019. His dissertation, “GPU Data Structures and Code Generation for Modeling, Simulation, and Visualization,” introduced methods that significantly accelerate and integrate key components of the virtual prototyping cycle, forming the basis for simulation-driven design tools with immediate feedback.
Design for real world engineering: integrating uncertainty into product assessment
Presentation Abstract
Traditional engineering design processes rely on fixed input values for loading conditions, geometry and material properties. Although this deterministic approach is straightforward, it does not embrace real world uncertainty and design approaches focus on very conservative worst-case scenarios or non-appropriate conditions. As a result, components may be overdesigned or may still fail when conditions differ from the assumed limits. To address this challenge, we have developed a design toolkit that incorporates uncertainty directly into computational analysis.
The toolkit integrates Monte Carlo simulation methods into the modelling process. Instead of using single values, statistical distributions are assigned to key inputs. The software then performs large numbers of simulations to produce a clear picture of how likely different performance outcomes are, allowing failure probabilities and sensitive parameters to be identified.
We demonstrated the toolkit using a sensor mounting bracket from a smart wearable. A traditional optimisation reduced the bracket’s weight by 30% while our probabilistic approach revealed that the design was tuned to an extremely unlikely drop event and still carried thermal failure risks. The output of our process provided additional detailed information to the design and engineering teams, enabling them to re-optimise and achieve a 50% weight reduction while maintaining an extremely low failure probability. This work highlights how probabilistic tools can support more reliable and efficient product development.
Speaker Bio
Greg is an engineering consultant with extensive experience in the development of consumer electronic products and specialist expertise in digital twin technologies and advanced simulation workflows. His work centres on combining physics based modelling with data driven methods to enhance product reliability, optimise performance and support informed decision making throughout the development cycle. He has significant experience in finite element analysis, predictive maintenance and the development of automated, reproducible pipelines for complex engineering computations.
Digitizing Body-in-White Development with MAS Synera
Presentation Abstract
The automotive industry is undergoing a profound transformation as digitals tools, automation, and computational design reshape traditional engineering processes. Within this context, Body-in-White (BIW) development with Synera MAS presents a major opportunity to modernize workflows, accelerate iterations, and improve both efficiency and sustainability.
Speaker Bio
Tilman Steininger is a Customer Success Manager for Agentic Engineering at Synera, driving scalable automation and measurable impact in complex engineering organizations. With over 20 years of experience across manufacturing, CAE-driven development, and digital transformation, he combines deep technical expertise with a strong focus on operational performance and ROI. His work centers on connecting CAD, CAE, and manufacturing processes to unlock structural efficiency and enable next-level engineering productivity.
Juan De Dios Escribano Felguera is a Team Lead for Body‑in‑White and Corrosion Development across all SEAT/CUPRA models, bringing extensive experience in automotive engineering, product development, and BIW industrialization. He focuses on modernizing engineering workflows through digital methods, automation, and computational design to accelerate structural development and enhance efficiency across the organization, working in close collaboration with Synera and multidisciplinary SEAT/CUPRA teams.
Large Engineering Models: Reimagining Design, Simulation, and Manufacturing.
Presentation Abstract
Engineering is entering a new paradigm where AI is no longer limited to accelerating isolated simulations, but is becoming a foundational layer of the industrial design and manufacturing stack. Across sectors such as automotive, energy, semiconductors, and aerospace, engineering workflows remain constrained by slow simulation feedback loops, fragmented CAD-to-CAE pipelines, and computational bottlenecks that limit design exploration and innovation speed.
In this talk, we present the vision of Large Engineering Models developed at Emmi AI: physics-aware foundation models designed to operate directly on industrial geometry and process inputs while replacing large parts of the traditional numerical simulation workflow. Rather than focusing on narrow surrogate models, this approach consolidates geometry processing, physics prediction, and post-processing into a unified AI-native engineering interface that provides near real-time feedback to engineers and designers.
As a concrete industry showcase, we present NeuralMould, our first commercially deployed Large Engineering Model for injection moulding. The model enables engineers to evaluate design changes such as gate placement, material selection, and process parameters with near-instant feedback, eliminating traditional meshing bottlenecks and dramatically accelerating design iteration cycles. In addition, we introduce Noether, our recently released open-source framework for building and deploying engineering AI models at scale. Noether provides the infrastructure for training, fine-tuning, and operating Large Engineering Models across domains, enabling industrial partners and researchers to develop their own AI-native simulation capabilities.
Speaker Bio
Quercus Hernández is a Senior Research Engineer at Emmi AI working at the intersection of computational mechanics, numerical simulation, and machine learning for engineering applications. His work focuses on developing scalable data-driven modeling approaches that accelerate high-fidelity physics simulations while preserving predictive accuracy. He has experience in multi-physics modeling, scientific computing, and AI-based surrogate modeling for industrial manufacturing processes. His recent work explores transformer-based architectures for large-scale engineering simulations and their application to real-world design and optimization challenges.
Organization:
Tech Soft 3D
Presenter:
Luis Salazar Betancourt, Ph.D.
HOOPS AI: Correlating CAD Geometry with Manufacturing and Business Process Information
Presentation Abstract
Advances in computational design and additive manufacturing have enabled increasingly complex geometry, yet industrial AI adoption remains constrained by a fundamental challenge: relating CAD design data to manufacturing behavior and downstream business processes. Geometry, process data, and enterprise information are typically analyzed in isolation, making it difficult to understand how design decisions propagate across the product lifecycle.
HOOPS AI introduces a technical framework that enables such correlation by transforming CAD and manufacturing data into unified, AI-ready representations. Built on the HOOPS platform, the approach focuses on extracting stable geometric and feature-level abstractions that can act as a common reference across engineering and non-engineering domains, without reliance on native CAD kernels or proprietary data models.
These representations enable systematic comparison, grouping, and retrieval of parts based on geometric and functional characteristics. Because they are shared across workflows, the same representations can be associated with manufacturing signals such as process variability, quality indicators, or production constraints and extended to link with business process information in a controlled and IP-safe manner. This makes it possible for AI systems to reason about relationships between CAD geometry, manufacturing outcomes, and operational drivers.
This presentation will focus on the architectural principles behind HOOPS AI and illustrate how unified representations support scalable integration of AI into industrial CAD and additive manufacturing workflows. Emphasis is placed on system design and interoperability rather than on specific models or algorithms.
Speaker Bio
Luis Salazar Betancourt, Ph.D. is a senior technologist at Tech Soft 3D, where he helps drive the company’s work at the intersection of 3D engineering, artificial intelligence, and machine learning. With a background rooted in computer-aided engineering (CAE) development, Luis has been instrumental in evolving Tech Soft 3D’s approach to intelligent engineering solutions and has played a critical role in the development of HOOPS AI, the company’s latest SDK.
He holds a Ph.D. in Computational Mechanics and Materials from Mines Paris and brings over 10 years of experience spanning computational mechanics, materials science, and high-performance software engineering. Known for his ability to bridge deep engineering expertise with product vision, Luis focuses on integrating complex 3D engineering data into modern AI workflows, enabling organizations to extract domain-specific intelligence through data-driven technologies.
Beyond the Specialist: How Istari + AI Expands Who Gets to Drive Innovation in Hardware
Presentation Abstract
AI coding tools didn’t just make developers faster, they changed who gets to build. A similar shift is underway for hardware.
This talk explores how AI paired with connected engineering infrastructure is lowering the barrier to entry for hardware innovation – not by replacing deep expertise, but by expanding the aperture of who can contribute meaningfully to the design process. When more of the right people can engage earlier, ideas move faster and better solutions emerge from unexpected places.
Speaker Bio
Rebeka “Cam” Melber is the Director of Programs at Istari, where she oversees program execution and works closely with government and industry teams adopting next-generation engineering capabilities. Her career has focused on helping complex defense programs transition to modern digital infrastructure and accelerating the adoption of new technologies within sensitive mission environments. Drawing on experience across the Department of Defense, her work centers on making advanced engineering tools accessible to the people who need them most. Rebeka has led several first-of-their-kind capabilities and brings a blend of strategic vision, engineering discipline, and operational execution to advancing how complex systems are designed and delivered.
You have an infrastructure problem, not an AI problem
Presentation Abstract
For engineering teams, AI excels at individual tasks like SQL queries, graphs and summaries. But, higher value work such as root cause analysis, physics-based ML and production deployment remains technically out of reach.
The common diagnosis is that the AI models need to improve. My experience suggests the opposite: the models are already sufficient, but the infrastructure surrounding them is not.
Using a real rotor balancing case with 50,000 rotors per year, I will show what it takes to move AI from task-level support to autonomous engineering investigation. Coupled with the right infrastructure, a model that had struggled on summary data uncovered a discriminating signal in 100 kHz vibration data that decades of domain expertise had overlooked. It then independently designed a deployment plan that reduced total balancing runs by 41%.
In this talk, I will walk through what was built, the output, and what it means for any organisation sitting on complex engineering data.
Speaker Bio
Javi is a Lead Solutions Engineer at Quix, where he helps customers get the most out of their hardware data. He was previously a senior data scientist at Orange, developing churn prediction, marketing mix modelling, propensity to purchase models and more. Javi is a Master’s lecturer and speaker specialising in pragmatic data science, causality and real time.
Immersive engineering: Navigating harnessing in true-scale 3D
Presentation Abstract
Spatial computing is often framed as a future breakthrough, but it’s already delivering irrefutable value, just not always where people expect. This session explores where immersive technologies are making a tangible impact in product development today, particularly in workflows like cable harnessing and automotive clay modeling, where teams still rely on physical prototypes to understand critical decisions.
Attendees will learn how spatial tools enable earlier, more confident decision-making by improving spatial understanding and reducing costly physical iteration. The talk will also address the practical realities of building for spatial hardware, from performance constraints to interaction design, offering a grounded perspective on what it takes to make these systems work in real-world engineering environments.
Speaker Bio
Oluwaseyi Sosanya is a design engineer and the CEO of Gravity Sketch; a London-based startup that is building an immersive 3D workspace built for ideation and collaborative problem-solving. Over his career, he has focused on challenging today’s traditional digital tools to develop user-friendly sustainable design solutions. As the CEO of Gravity Sketch, he is pursuing the goal of making emerging technologies accessible across the product development process, and lowering barriers to entry with human-centric user experiences.
Computational Color and Texture for Additive Manufacturing
Presentation Abstract
While we continue to push the boundaries of computational geometry, the integration of color and texture represents a vital and under-explored frontier in additive design. As multi-material hardware and sophisticated file formats become increasingly accessible, the ability to program color and texture directly into the design workflow emerges as a powerful new tool set. This session explores our experiences building up such a tool set and will dive into effects that traditional modeling tools cannot easily capture. Our talk will include:
• Optical Phenomena: Implementing shimmering and view-dependent effects.
• Embedded Data: Integrating “hidden” messages and encoded information within a part’s skin.
• Procedural Topography: Using color-driven surface displacement.
• Gradient Control: Managing complex geometry-driven color and texture transitions.
• Realistic In-workflow Rendering: Displaying accurate as-printed colors and textures within the design workflow.
• Procedural Customization: Automating the generation of unique colorways and textures across product families, including the use of AI-generated content.
We will examine the workflows required to translate digital intent into reliable physical results, while highlighting common pitfalls and how to navigate them. To bridge the gap between pixels and atoms, we will pass around physical samples for a hands-on look at these techniques.
Speaker Bio
Mary Baker is the proprietor of Palace3D, a consulting and contracting firm for computational design for additive. From 2018-2025, she was an Architect in Computational Design for Additive Manufacturing at HP Inc. where she helped introduce procedural design tools such as Houdini to the additive design community. She also co-led the automated design of 3D-printed molded fiber tooling parts for HP’s sustainable packaging business, wrote the back-end automation engine for turning AI-generated content into 3D-printable customized products, designed HP’s 316 Stainless 3D printed metal jewelry, and built a variety of software tools to help make additive design for color more intuitive and reliable. Before HP she was on the faculty of the CS and EE departments at Stanford University where she graduated 7 PhD students.
Personalized Proprioception: Automated Infill Generation for Gradient Compressibility
Presentation Abstract
This presentation will discuss the creation of the Sequence Dynamic Insole Generator, an automated and bespoke tool to generate custom additively manufactured insoles based upon human biomechanics data. This process was built using SideFX Houdini to leverage proceduralism and automated data handling in order to create a robust, operator-controlled process for the creation of individual biomechanic insoles. Rather than creating 3D CAD Geometry and relying on an off-the-shelf slicer, this workflow generates a variable density infill toolpath pattern based on input parameters and geometry.
Custom pathfinding algorithms were developed for print optimization and manufacturability, embedding constraints and print parameters directly into G-Code Generation. Gradient compressibility and energy absorption enable the insoles to modulate proprioceptive input across the body’s kinematic chain, supporting adaptive changes in the user’s dynamic movement patterns. We will discuss algorithmic strategies for toolpath generation, automated adherence to printing parameters, and integration with digital fabrication methods, demonstrating how procedural design environments can bridge biomechanics, computation, and manufacturing in next-generation product creation systems.
Speaker Bio
David Burpee is a multidisciplinary Computational Design Leader based in the Pacific Northwest, with expertise spanning Footwear, Apparel, Consumer Goods, Automotive, Medical, and Architecture industries. He lectures on Computational Design and Algorithmic Thinking at the University of Washington and is a Computational Researcher on a National Science Foundation grant exploring Engineered Living Materials (ELMs).
With over a decade of Computational Design experience, David has delivered advanced design strategy, tools, and training for companies including Nike, PUMA, FILA, General Motors, Harry’s Razors, and EQLZ. His work demonstrates a proven methodology that merges creativity, deep technical capabilities, and broad market impact.
Originally trained as an Architectural Designer with a Master of Architecture from USC, David has contributed to highrise and supertall projects in Los Angeles, Seattle, and across Asia. His work integrates computational approaches at every scale, from skyscrapers to small installations.
Driven by a passion for biomimicry, generative systems, and sustainable innovation, David applies computational design to address complex ecological and social challenges through creative, high-performance solutions.
Keynote Presentation
Organization:
Nike
Presenter:
Yuan Mu
Textile, Rewritten
Presentation Abstract
Nike’s recent developments in computational fabrication and materials engineering have catalyzed a transformation in textile production, shifting from analog processes toward digitally controlled, additive and hybrid manufacturing techniques.
This talk examines the emergence of 3D-printed and algorithmically designed textiles through as a new design language for high performance products across footwear, apparel, and accessories. By embedding athlete motion data into parametric workflows, designers can rapidly iterate and precisely control optimized structural patterns and material deposition to create seamless textile matrix, reducing material waste and assembly steps.
Speaker Bio
Based in Portland with a penchant for travel, Yuan Mu currently works as a computational designer in Nike Innovation. By day, she specializes in designing and researching wearable technology, particularly focusing on innovative methods of make. In her spare time, she explores the intricacies of digital design and fabrication tools, expands her skills and perception through a variety of professional experiences. Prior to joining Nike, she worked in architectural and 3d visualization offices in Los Angeles.
A New Ecological Simulation Framework for Rhino/Grasshopper
Presentation Abstract
Rhino.Ecologic® is an advanced ecological simulation framework built for Rhino and Grasshopper, designed specifically for landscape architects and urban planners working with computational design. It allows users to seamlessly integrate ecological simulations into their existing 3D design workflows, supporting nature-inclusive strategies at both large and small scales.
With Rhino.Ecologic®, designers can generate location- and time-specific 3D species distribution maps, along with biomass and biodiversity simulations, enabling data-driven ecological decision-making throughout the design process.
With this presentation, you will get a concise overview of the new Rhino/Grasshopper plugin, an introduction to ecological simulation in AEC, and a demo showcasing how the tool works in practice.
Speaker Bio
Verena Vogler leads the R&D team and the programming curriculum at McNeel Europe. She holds a Diploma (Leipzig University of Applied Sciences) and a Master’s in Architecture and Computational Design (IaaC), and earned her doctorate in Engineering from the Chair of Computer Science in Architecture at Bauhaus University Weimar.
Organization:
AiA Life Designers
Presenter:
Michele Pescatore
Empowering Architects with Early-Stage Environmental Intelligence
Presentation Abstract
AiA Life Designers is a French multidisciplinary practice of architects and engineers, bringing together multiple professional areas of expertise and exploring the potential of digital and computational design approaches for several years. From serial design strategies and complex infrastructural structures to solution optimization, application development, interoperability workflows, data management, and algorithmic approaches, the field of possibilities is vast. Within this broad digital landscape, environmental performance has emerged as a critical focus, particularly at the earliest stages of architectural design. Architectural decisions made in the earliest stages of architectural design have the greatest influence on a project’s environmental performance, yet architects often work without timely or actionable feedback. Existing environmental analysis tools are powerful but poorly suited to early-stage design decision making: they require highly detailed BIM models or expert-level parametrization. As a result, environmental studies quickly become outdated and no longer reflect the evolving state of the project.
To bridge this gap, we developed a data-driven, simulation-driven workflow implemented through a streamlined Revit plug-in that allows architects to perform real-time environmental assessments directly within their everyday design environment. Built with Rhino.Inside.Revit, Grasshopper, and the open-source Honeybee and Ladybug ecosystems, the tool leverages parametric design and design automation to deliver simplified but meaningful indicators such as envelope performance checks, site permeability, carbon impact estimates, and solar and daylight analyses. This approach enables architects to rapidly test assumptions, adjust parameters, and immediately visualize environmental implications through performance-based design feedback, well before formal engineering studies begin. By embedding integrated design workflows into standard architectural practice, the tool supports rapid iteration without disrupting existing design processes. By making environmental intelligence accessible, intuitive, and embedded within daily workflows, the solution fosters a culture of eco-design and positions architects as proactive contributors to environmental quality from the very first sketches. In this presentation, we will demonstrate several real-world project case studies, share concrete outcomes, and discuss the scalability of computational design tools across a large architectural organization.
Speaker Bio
Michele Pescatore is an architect and BIM referent with a strong passion for computational design and emerging technologies. Trained and professionally active in both Italy and France, he has developed solid expertise across multiple architectural domains, with a particular focus on BIM with Revit, computational design, and digital fabrication using Grasshopper and Rhinoceros.
He has worked on international competitions and construction projects of various scales and phases, applying technology as a driver for innovation and design quality. Proactive, detail-oriented, and an effective team collaborator, he is committed to continuous learning and to advancing the integration of digital tools into everyday architectural practice.
Carol Fanjul is a BIM Manager and licensed Architect at AIA Life Designers in Paris, with over 10 years of experience in architecture and engineering, including nearly 8 years in BIM management. Her work focuses on enhancing project performance through the development of BIM standards, workflows, and digital tools.
She is involved in the coordination of complex BIM projects and the implementation of collaborative processes across disciplines. She is particularly interested in the integration of data-driven methods and environmental intelligence into early-stage design.
Her approach combines technical rigor and a strong focus on usability, with the aim of making digital tools more efficient and accessible for design teams.
Organization:
University of Stuttgart, IntCDC
Presenter:
Claudia Valverde
Architected Porosity Informed by Real-World Data for More-Than-Human Thermal Comfort
Presentation Abstract
This presentation introduces a design and research framework that integrates geometry generation with real-world climatic and ecological data to support more-than-human thermal comfort in the exterior of building envelopes. Over the past two years, I have developed architected porous cellular structures—periodic and non-periodic—based on adaptive density minimal surfaces (ADMS) and triply periodic minimal surfaces (TPMS). These structures serve as protective envelopes for nesting tubes used by cavity-nesting wild bees.
The novelty of this work lies not in the digital modelling itself, but in demonstrating how pore size and spatial gradients can be tuned to buffer heat threats inside nesting cavities, and how these porous morphologies behave under real outdoor conditions. Full-scale and small-scale prototypes were installed on real building settings, where they were exposed to solar radiation, diurnal temperature swings, summer heat events and varying humidity. Continuous monitoring revealed how these structures process and respond to environmental information—delaying heat peaks, modulating temperature transfer, and interacting with passive evaporative cooling strategies.
In parallel, wild bee occupation of the prototypes provided biological feedback, confirming which geometries are perceived as suitable nesting habitats. Bringing together digital modelling, outdoor performance testing and ecological observation, this research proposes a design approach that situates building envelopes as active interfaces capable of supporting more-than-human thermal comfort in urban environments.
Speaker Bio
Claudia Valverde is an architect and researcher exploring more-than-human thermal comfort and bioreceptivity through architected geometry and material performance. She holds a Master in Industrial Design for Architecture from the Politecnico di Milano and is currently a PhD candidate and Research Associate at the University of Stuttgart’s Cluster of Excellence IntCDC (ITKE). Her work investigates how porous morphologies, additive manufacturing and façade-integrated systems can support non-human species—particularly cavity-nesting wild bees—within real outdoor building conditions. Through generative design methods, fabrication experiments and environmental monitoring, she develops prototypes that expand the ecological role of architectural surfaces and position geometry as a mediator between climate, materials and living organisms.
Computational Design of Personalized CPAP masks
Presentation Abstract
CPAP therapy is used to treat patients with sleep apnea by providing constant air pressure through a mask during sleep. For many patients (75,4%), standard masks fit poorly or require tight adjustment, causing issues such as pressure points or leakage resulting in dry eyes, skin irritation, and finally even therapy discontinuation (30%). Personalized CPAP masks can improve fit and comfort and reduce leakage.
However, early designs face challenges due to the interaction between soft facial tissue and mask materials. During sleep, facial deformation and mask compliance vary with posture, often leading to leakage. In this study, we analyzed 3D scans of individuals lying in multiple positions to model how posture affects mask shape. To improve further, a sensor instrument was developed to measure pressure points while wearing existing masks, this data was used to optimize a computational model that integrates head posture, 3D scan data, and facial softness to guide mask personalization.
Two mask concepts were developed: (1) a fully personalized mask with a 3D-printed soft interface and a rigid hard part available in three sizes, and (2) an intermediate personalized layer that fits between a standard mask and the face to improve fit without replacing the entire mask. Both designs were optimized for 3D printing with Lynxter silicone -improving slicing, print time, neatness, and thickness distribution – as well as for manufacturing with SLA cocoon molding and silicone casting.
Clinical testing in a dutch hospital is planned for March to evaluate the fit of both designs. Results are expected to provide insights into personalized mask design, the feasibility of a 3D scan–model–print workflow in a regional hospital, and the potential of 3D-printed masks for tailored CPAP therapy. When this strategy for generating the model and manufacturing proves potential, the strategy can be extended to other respiratory masks.
Speaker Bio
Anne Pasman studied Industrial Design with a focus on Emerging Technology Design. She currently works at the Industrial Design research group on a line exploring new technology in relation to design. With research areas combining AM and computational design for both personalized medical products and experimenting with innovative materials such as wood, concrete, and silicones.
Emmy Kerssen studied Industrial Product Design. Currently she is working at the Industrial Design Research Group of Saxion University of Applied Sciences on a line exploring new technology in relation to design. Her work focuses on computational for personalized medical products and additive manufacturing of new materials like foam and silicones. In addition, she works at FabLab Enschede where she works as a designer, where she also educates students on the use of digital fabrication technologies.
Organization:
Raven
Presenter:
Moritz Rietschel
Raven AI: The future is in the spaghetti
Presentation Abstract
Your most valuable design tools already exist. But they are buried under learning curves, scattered documentation, and years of institutional knowledge or old forum posts. What if AI didn’t replace any of that but made it accessible?
Raven is a code agent for Grasshopper that works across 900+ existing plugins through the native API. We think of it less as a product and more as infrastructure: a layer that lets someone access 20 years of community built tools, decades of firm knowledge, and lets long-time Grasshopper users discover features in their own tools they didn’t know they had. The tools and knowledge was always there, you just couldn’t get to it.
This talk is about what it takes to build AI that works with CAD. How we squeeze 3D geometry and graph logic into a 1D token stream. How we fight the LLMs assumptions when Rhino disagrees. Why do reasoning models sometimes spiral into nothing useful? We show our guardrails for keeping reasoning models on track, how we draw from scattered public knowledge and your three person team alike, and what it means to design for recombination when every project needs a unique assembly of plugins and every user has a globally unique Grasshopper setup.
We’ll share what breaks, what surprised us, and how we are thinking about AI within creative ecosystems rather than instead of them.
Speaker Bio
Moritz Rietschel is co-founder of Raven, building at the frontier of AI capabilities and CAD. His work builds on research at UC Berkeley from early 2024, bringing early LLMs into CAD as AI-collaborative tools with Prof. Kyle Steinfeld and robotic fabrication with Prof. Simon Schleicher. Moritz has lectured internationally on AI-based creative tools, published at ACADIA and the IJAC journal, hosted workshops at MIT and Tongji and contributed to international research conferences.
Organization:
SimScale
Presenter:
Jon Wilde
From Tools to Agents: How Agentic Engineering Workflows Are Reshaping Simulation-Driven Product Development
Presentation Abstract
Simulation is central to engineering decision-making, yet in many organizations it remains an expert-driven activity rather than a scalable capability embedded across new product introduction (NPI). As product complexity grows and timelines compress, the key challenge shifts from solver accuracy to workflow coordination: when to simulate, at what fidelity, and how results inform decisions.
This talk introduces agentic engineering workflows — AI-driven systems that guide simulation tasks, recommend appropriate model fidelity, and support interpretation while remaining grounded in validated physics. By combining Engineering AI with Physics AI, these workflows move beyond static automation toward context-aware orchestration of design and validation processes.
Examples include AI assistants that assess simulation readiness — reviewing mesh quality, boundary conditions, and convergence behavior — and CAD-triggered workflows that initiate physics-based validation and surface performance trade-offs. Engineers remain in control, with AI operating within defined guardrails.
Agentic workflows enable earlier and broader use of simulation while preserving traceability, verification standards, and domain expertise. Rather than replacing traditional CAE, they represent an evolutionary step in how simulation insight is generated, reused, and governed across the product lifecycle.
Speaker Bio
Jon Widle is Vice President of Product at SimScale, where he leads the development of cloud-native simulation, Engineering AI, and Physics AI solutions for industrial engineering teams.
Redefining mechanical engineering in the age of AI
Presentation Abstract
Mechanical engineering has been limited by the capabilities of traditional CAD software, most of which are built on architectures that are more than 30 years old. How can we design the products of the next 30 years using technology created three decades ago?
AI can introduce a real paradigm shift in how we conceive and develop products, but engineers must remain at the centre of the process. As an engineer, I know that the workflow is often more important than the final result. A great result coming from a black box is never truly great.
In this talk, I will discuss the changing role of mechanical engineering in the age of AI, the key bottlenecks slowing down AI adoption, and practical ways to integrate AI into engineering workflows in a sustainable way for the aerospace and automotive industries, where standards and certification requirements are highly restrictive.
Speaker Bio
Rhushik Matroja is the CEO and Co-Founder of Cognitive Design Systems, a deep-tech company building next-generation software for generative design and manufacturability intelligence. With more than a decade of experience in mechanical engineering, additive manufacturing, and simulation, he focuses on reducing design cycle time for aerospace, automotive, and industrial sectors. His work brings advanced automation, performance optimisation, and design exploration into everyday engineering workflows, helping companies accelerate product development and make better decisions.
Organization:
ELISAVA-DIGIFAB-IAM3DHUB
Presenter:
David Adrover Iglesias
CD&AM Ecosystem in Barcelona, from university to industry
Presentation Abstract
Barcelona is a leading hub for additive manufacturing and computational design, with a strong ecosystem of companies, institutions, and initiatives. David Adrover will present how key organizations collaborate to develop talent, build skills, and drive innovation in these fields. This includes academic training through the Master in Computational Design and Additive Manufacturing at ELISAVA, as well as support for start-ups and industry projects through initiatives such as DIGIFAB and IAM3DHUB. The talk will showcase successful examples emerging from this collaborative network and highlight opportunities to connect, collaborate, and innovate in Barcelona’s growing CD & AM ecosystem. If you want to learn more about to collaborate and impulse innovation together in Barcelona in CD & AM don’t miss the talk
Speaker Bio
David Adrover is a Mechanical Engineer specialized in Computational Design and Digital Manufacturing. He serves as General Secretary of IAM3DHUB (International Advanced Manufacturing 3D Hub), operated by the technology center LEITAT. He is also President of Addliance – The European AM Hubs Alliance, and Managing Director of the DIGIFAB network, part of the Catalan Society of Technology, where he is a Board Member.
In addition to his leadership roles, he collaborates with several academic institutions, including ELISAVA Design and Engineering University, the Polytechnic University of Catalonia (UPC), and ESADE Business School, where he teaches topics related to Computational Design and Additive Manufacturing.
Bridging Data to Geometry with Implicit Modeling
Presentation Abstract
Engineering design is increasingly shaped by diverse sources of data (e.g. scans, images, measurements, simulation, requirements, reference geometries etc.). Traditional geometry modeling approaches often struggle to integrate these datasets in a flexible and scalable way. Implicit modeling offers a powerful alternative: a geometry representation that naturally incorporates engineering data into the modeling process as spatially varying fields capable of driving local and global design behavior.
This presentation explores how Simcenter Inspire leverages implicit modeling to create robust, data-driven design workflows. Showcasing methods for linking complex data inputs directly to geometric parameters, enabling both automated design optimization and fine-grained user control.
Speaker Bio
Wesley Essink leads the Implicit and PolyNURBS modeling capabilities used in Simcenter Inspire at Siemens. His role focuses on developing Inspire’s non-traditional modeling capabilities enabling users to create complex, simulation-driven design workflows that are both intuitive and efficient for engineers and designers. He was the co-founder and CTO of Gen3D which was acquired by Altair in 2022 and subsequently acquired by Siemens in 2025 and has over a decade of experience of academic and industrial R&D in design and manufacturing.
An Engineer’s Approach to Integrating Machine Learning in Generative Design Tools
Presentation Abstract
Machine learning and Artificial Intelligence have enormous potential as tools for generative design in engineering. However, most industry efforts remain stuck in research prototypes, brittle bespoke models, or disconnected add-ons that rarely survive real engineering workflows. In this talk, we will present an engineer’s approach to integrating machine learning directly into production-ready generative design tools, drawing on our experience building the fastest physics-driven thermo-fluid optimization platform on the market. Rather than replacing physics with opaque black boxes, our methodology uses ML only where it strengthens engineering outcomes.
I will show how ToffeeX’s ML developments accelerate design exploration and automation while preserving full control of engineering intent, seamlessly extending our existing topology optimization engine which is already used daily in real production environments. This talk highlights why our approach, built on smart algorithmic design rather than brute-force model training, achieves the reliability, manufacturability, and speed required for real-world engineering.
Speaker Bio
Thomas is an aerospace engineer specialized in fluid dynamics, heat transfer, and optimization. He holds MEng, MRes, and PhD degrees from the Department of Aeronautics at Imperial College London and currently leads R&D efforts at ToffeeX. Known for combining rigor with curiousity and a sense of humour (he once tried to cook a steak in a hypersonic wind tunnel — it had a good sear, but was cooked blue), he has nearly a decade of experience building CAE tools and engineering processes.
Conjugate Heat Transfer Optimization for Turbine Blade Thermal Performance Using Field-Driven Design
Presentation Abstract
Turbine blade thermal management demands tight coupling between design exploration and high-fidelity simulation—yet traditional workflows separate these domains, limiting the design space that can be practically explored.
We present an integrated parametric optimization framework enabled by nTop’s robust implicit geometry engine. Field-driven design representation eliminates mesh regeneration between design variants, while a Lattice Boltzmann Method formulation for conjugate heat transfer removes the need for explicit fluid-solid interface handling. This architectural unity—implicit geometry paired with interface-agnostic thermal transport—permits fully automated design-simulate-optimize loops on complex internal cooling geometries.
The GPU-native solver has been validated against finite-volume baselines on canonical heat sink geometries, demonstrating peak temperature agreement within 0.5% while achieving approximately 200x reduction in time-to-solution on consumer-grade hardware. These evaluation times make high-fidelity CHT practical as an inner-loop optimization objective rather than a final verification step.
We demonstrate the workflow on turbine blade internal cooling channels, where parametric control over fin positioning drives systematic exploration of the thermal-structural design space. Results show automated identification of Pareto-optimal configurations balancing thermal performance, pressure drop, and additive manufacturing constraints.
This collaboration between nTop and Siemens Energy bridges computational design research with industrial turbomachinery requirements.
Speaker Bio
Organization:
Rafinex
Presenter:
André A.R. Wilmes
Rafinex
Presentation Abstract
Speaker Bio
NeuralShipper: Generative AI for the Next Generation of Ship Design and Manufacturing
Presentation Abstract
Water transport, which accounts for approximately 90% of global trade, is essential to economic growth but poses a significant environmental challenge. In 2020, shipping emitted 1.4 billion tonnes of CO₂, nearly 3% of global emissions. Without decisive action, this figure could rise to 18% by 2050. To address this, the International Maritime Organization (IMO) is increasing regulatory and financial pressure on industry stakeholders to cut emissions by at least 40% by 2030 through the adoption of innovative technologies.
Achieving these targets requires optimising vessel performance from design through to operation. In fact, 80% of a product’s environmental impact is determined at the design stage, making it the most effective point for influencing a ship’s environmental footprint. Experienced shipbuilders recognise that design and engineering decisions made at this stage affect 85% of total construction costs and approximately 90% of overall vessel performance.
However, the maritime industry is often perceived as conservative compared to other transport sectors such as automotive and aerospace. Existing design practices are not suited to developing tools that enable true innovation.
This is where Compute Maritime enters the picture, offering AI-powered design tools and data-driven solutions for maritime sustainability.
Our flagship product, NeuralShipper, is the world’s first generative AI co-pilot for the design, optimisation, and simulation of maritime systems. It can handle everything from ship hulls and propellers to hydrofoils and rudders, all within a single platform.
Unlike conventional tools, which are often restricted to specific ship types, NeuralShipper is universally adaptable. Whether naval architects and marine engineers are working on a fuel-efficient cargo ship, a specialised workboat, a sleek yacht, or a complex naval vessel, NeuralShipper is designed to support them and accelerate the maritime industry’s progress towards net-zero emissions.
With only a few design specifications as input, NeuralShipper can generate thousands of tailored design concepts within minutes, each meeting the specified criteria. This eliminates the need for manually drafting preliminary sketches or parametric CAD models, enabling designers to evaluate optimised options in real time. Users can also define custom constraints and performance criteria to create bespoke solutions.
Traditional design tools rely heavily on user expertise and deep familiarity with complex software, making the process slow and innovation-limiting. In contrast, NeuralShipper
streamlines the entire process. It acts as a collaborative AI designer, working alongside human experts to develop solutions that are both performance-efficient and innovative. This significantly accelerates concept development and allows teams to move swiftly into detailed design without compromising on creativity, innovation, or sustainability.
At the core of NeuralShipper lies our Large Geometric Foundation Model, trained on over 100,000 ship designs encompassing nearly every type, shape, and category. This extensive dataset gives NeuralShipper a unique ability to generate solutions that would be difficult or even impossible for human designers to achieve when working with incomplete inputs or domain-limited knowledge. Such scenarios are common in cutting-edge projects exploring next-generation propulsion systems and alternative fuels.
Importantly, NeuralShipper is the first generative AI model capable of directly outputting a CAD model, primarily in the form of a NURBS surface. One of the critical challenges facing existing 3D foundational models is surface quality. These models, which often rely on low-level shape representations, frequently fail to capture the geometric precision required for reliable performance analysis. In engineering contexts, where even minor surface imperfections can significantly influence outcomes, ensuring smoothness and validity is crucial.
Addressing this challenge, and achieving high surface quality and physics-informed accuracy, has been central to NeuralShipper’s innovation.
Speaker Bio
Shahroz is the founder and CEO of Compute Maritime, where he and his team are developing next-generation intelligent design frameworks for the maritime industry. Their work aims to radically transform how vessels are designed, simulated, optimised, and manufactured.
Prior to this, Shahroz served as Research and Development Lead at BAR Technologies, a company founded by the former Group CEO of Aston Martin and McLaren Formula One. He has also been the founder of two successful start-ups.
Shahroz has been an early adopter of artificial intelligence in the ship and yacht design industry. One of his pioneering innovations was ModiYacht—the first patented tool for semantic and attribute modelling of yachts—which he developed in 2015 while working at CAD Smart Lab as a TÜBİTAK Fellow. This tool is currently used by Samsung and other major maritime companies.
Following this, he founded Barinlabs Group to integrate generative AI into product design. This initiative led to the launch (and eventual acquisition) of his first start-up, even before its official product release. He subsequently pursued a series of research projects focused on transforming what is often regarded as one of the world’s oldest and most conservative industries: maritime.
Among his most notable inventions is ShipHullGAN, the world’s first generative foundational model for ship design and optimisation. He developed this model during his PhD in 2023 as a Marie Skłodowska-Curie Fellow—one of Europe’s most prestigious research fellowships. Funded by the European Union, the project received approximately €3 million to advance deep learning in engineering applications.
ShipHullGAN attracted significant attention from industry leaders such as BAE Systems, a major British defence contractor, as well as other prominent European companies. Shahroz eventually founded Compute Maritime and launched a commercial variant of ShipHullGAN known as NeuralShipper.
Beyond Surrogates: Foundational AI for Physics-Native Design
Presentation Abstract
A new generation of AI models is emerging — not just faster approximators, but intelligent systems that understand and generate physics.
In this talk, Alan Patterson, CEO of BeyondMath, will explore how foundational physics AI breaks from the surrogate modeling paradigm. Unlike models that rely on customer-provided simulation data or narrow datasets, BeyondMath’s models are trained on self-generated data rooted in first principles — not interpolating outcomes, but learning the physical laws and structure of the design space itself.
This approach enables something radically new: generalizable, physics-consistent predictions at near-CFD fidelity, delivered in seconds — and without the need to retrain when a geometry changes. It opens the door to simulation-native design workflows, where simulation is not a bottleneck but a continuous, integrated part of ideation and optimization.
Alan will share:
– Why surrogate AI models struggle in real-world engineering
– What it means to build a foundational model that learns physics, not data correlations
– Case studies from sectors like motorsport and energy
– How these models enable new kinds of design tools and thinking
This is not an evolution of simulation — it’s a rethinking of how AI and physics interact. Foundational AI for physics is here, and it’s reshaping the very act of designing the physical world.
Speaker Bio
Wasil Rezk is the Chief Commercial Officer at BeyondMath, a company pioneering foundational AI that learns and generates physics, enabling real-time, physics-native design without relying on surrogate models or customer data. BeyondMath’s models are trained on first-principles-derived data, offering engineers high-fidelity simulation at unprecedented speed and scale.
Wasil previously worked at Palantir Technologies and AWS, where he specialized in opening new commercial operations across the United States, Europe, and the GCC, forging strategic partnerships and securing multiple enterprise contracts. He holds a BA in International Relations from Georgetown University.
Organization:
Generative Engineering
Presenter:
Laurence Cook
Beyond Parametric: Explorable Simulation for Real Design Iterations
Presentation Abstract
Design space exploration promises engineers the ability to evaluate thousands of variants in hours. The prerequisite—rarely discussed—is building a parametric geometry model that won’t break. This front-loaded investment assumes you know your design space before exploring it. Real engineering iteration isn’t parametric. Design changes are structural and topological: fundamentally different geometries evaluated against the same performance criteria.
Setup time, not simulation time, is the real bottleneck in enabling simulation-driven exploration for real engineering teams.
At Generative Engineering, this is what we are addressing. We enable explorable simulations that work with both parametric exploration and novel designs. We will demonstrate how this is unlocked by using existing simulation data inside engineering organisations to train geometry preparation models rather than physics surrogates.
This lets the messy, collaborative, real design iterations be simulated with zero setup cost. But the same principle extends further: if a model understands what makes geometry simulation-ready for a given analysis, it can generate that geometry directly. We’ll show work on closing this loop—from a sketch or annotation on a previous design to simulation-ready geometry, collapsing the traditional boundary between design intent and engineering validation.
Speaker Bio
Laurence Cook is Co-Founder and Chief Product Officer at Generative Engineering.
After academic work at MIT, Stanford, and Cambridge, Laurence brought his computational engineering to the commercial industry and is now leading the initial commercial product of Generative Engineering. A 20 strong team of computational and commercial experts who have currently raised €13m.
Real-time Multi-Physics Collaboration for Real-world Engineering
Presentation Abstract
Real multiphysics necessitates coordination along multiple dimensions: one is of course the mixed physics, another is the heterogeneous data types & modeling approaches, and a third is the multidisciplinary (mis)communication. AI shoved into engineering workflows does not eliminate these gaps, rather it exacerbates them. For instance, numerical solver outputs and ML data structures are incompatible, with no shared representations to bridge them. At the human level, CAE experts, Mech/Aero/Nuclear Engineers, Systems Engineers, and AI/ML Engineers all operate in different conceptual frameworks and thus lose valuable time and information at every handoff.
This talk presents Pasteur Labs’ approach to streamlining all three dimensions of multiphysics with the cohesive “Simulation Intelligence (SI) Platform”, emphasizing three plug-n-play products. SI Testbeds automate data generation end-to-end, from preprocessing through simulation runs and postprocessing, producing ML-ready CAE datasets at scale. SI Workspaces enable the flexible composition of next-generation multi-physics pipelines by stratifying surrogate models, optimization methods, and uncertainty quantification tools, with automatic differentiation as a first-class citizen throughout. Tesseracts are end-to-end differentiable containers that encapsulate heterogeneous numerical solvers and surrogate models behind a unified API, making otherwise incompatible components interoperable by design across software, hardware and teams. All prioritize modularity, scalability, and traceability, providing the engineering cohesion that is needed to adopt Physics-AI.
The SI Platform is elucidated by two different types of engineers working on two concrete engineering cases, with real-time collaboration: surrogate-based acceleration and optimization of centrifugal pump performance, and end-to-end gradient-based parametric optimization of rocket grid fins.
Speaker Bio
Nikolas Borrel-Jensen is a Simulation Intelligence Engineer at Pasteur Labs, where he develops approaches combining machine learning and simulation to create in-silico environments for human–machine collaboration across engineering domains such as thermal management, structural mechanics, fluid dynamics, acoustics and combustion.
Oliver Littlewood is a CAx Simulation Engineer at Pasteur Labs with a background primarily in motorsport aerodynamics, including taking time out of studies to work as a CFD Aerodynamicist in Formula 1. He is interested in all things CAE/optimisation with current work focusing on how legacy and future tools can merge to produce Simulation Intelligence
From Surrogates to Large Physics Models: Making AI-Native Engineering Work in Production
Presentation Abstract
Engineering is undergoing a fundamental shift. AI is no longer a point solution for accelerating isolated simulations — it is becoming a core layer of the engineering stack, reshaping how physical systems are designed, tested, and brought to production.
In this talk, Nico will share how PhysicsX is moving beyond narrow, task-specific surrogates toward Large Physics Models (LPMs). Using external aerodynamics as a concrete example, he will introduce PhysicsX’s latest model, trained on a large and diverse corpus of vehicle geometries, and show how it delivers real-time aerodynamic intelligence that generalizes to new concepts with minimal fine-tuning. Through production case studies, the session will demonstrate how AI-native workflows are compressing development cycles from months to weeks — enabling broader design-space exploration, continuous optimization, and faster, more confident engineering decisions.
Beyond aerodynamics, Nico will place this work in a broader AI-native engineering context. He will explore how Large Physics and Geometry Models, combined with adaptive workflows, connect design, simulation, and manufacturing into a seamless continuum. The result is not simply faster iteration, but a structural change in how engineering teams operate — where physics AI augments human judgment at scale, unlocking levels of performance and creativity that were previously out of reach.
Speaker Bio
Nico is the Co-Founder and Director of Engineering at PhysicsX, a pioneering deeptech company at the forefront of AI and engineering, dedicated to driving breakthrough innovations. Collaborating across Delivery, R&D, and Platform teams, he leads the development of advanced computer-aided engineering (CAE) and AI-accelerated design optimization methodologies. With extensive expertise in both engineering and deep learning, Nico works at the intersection of these fields to meet the important challenges of our time and build beyond human imagination. Prior to co-founding PhysicsX in late 2019, Nico built his career in the automotive and motorsport industries working for companies like Bentley Motors, Audi Motorsport and Mercedes-Benz.
Event Gallery
Venue
The symposium will take place at the Barcelona Biomedical Research Park (PRBB), a major international research center located on the Mediterranean coastline.
Carrer del Doctor Aiguader, 88, Ciutat Vella, 08003 Barcelona, Spain
The venue sits directly on the waterfront, with networking sessions held on the outdoor deck overlooking the ocean, offering space for informal discussion alongside the technical program.
