AEM Events

AEM Mechanics Seminar

Symmetry in Mechanics and AI: From Elasticity Classes to Equivariant Networks

Prof. Ryan Elliott
Department of Aerospace Engineering and Mechanics, University of Minnesota

A central challenge in modern AI for physical applications is embedding geometric structure directly into network representations. As frameworks like equivariant neural networks become standard, there is a growing need for general, algorithmic tools to characterize and manipulate invariant tensor spaces. Continuum mechanics offers a rich history and a mature mathematical toolkit for solving precisely this class of problems.

In this chalkboard lecture, we begin with the historical journey of material symmetry classification, detailing how a deeper group-theoretic perspective corrected decades of convention by reducing the number of elasticity classes from 10 to 8. We then present the general algebraic solution recently established by Olive and Auffray, demonstrating how harmonic decomposition paired with the "clips" operator (⊙) Provides a systematic procedure to algorithmically classify the symmetry of arbitrary tensor-based physical relationships. These geometric perspectives and formal classification tools offer a critical framework for the ongoing advancement of physical AI.

AIAA Study Night

Join the UMN chapter of the American Institute of Aeronautics and Astronautics for a study night.

AEM Colloquium Series: Professor Ran Dai

Smart Decision-Making for Autonomous Systems in Space Exploration Missions

Many autonomous systems benefit from efficient operations and advanced autonomy levels in ambitious space missions, from human exploration of Mars to on-orbit servicing, assembly, and manufacturing (OSAM) missions. However, complex system dynamics, changing operating environments, and stringent mission constraints make reliable, resource-efficient operations challenging. With limited human intervention, real-time decision-making that connects motion planning with task-level reasoning plays a critical role in assuring the reliability and performance of such a system toward mission success.

This talk will present our research on dynamic modeling, scalable optimization, and machine learning based optimal control methods that collectively contribute to advanced decision-making strategies for efficient autonomous systems in space exploration missions. The discussion will highlight applications in two distinct types of autonomous systems. This first concerns space vehicles in complex missions involving multiphase or hybrid operations where onboard propellant is limited and timely ground support is unavailable. The second focuses on origami-inspired deployable systems for OSAM, where systems automatically adapt their shape/functionality to mission needs. These applications share a central challenge: translating complex dynamics and mission requirements into decisions that can be computed and implemented in real time. The talk will highlight how advances in modeling, optimization, and learning are leveraged to support both precise spacecraft motion and adaptable robotic configurations. The overarching goal is to enable autonomous space systems to accomplish increasingly complex missions while making effective use of limited onboard resources.

Dr. Ran Dai is a professor in the School of Aeronautics and Astronautics at Purdue University. Before joining Purdue, she was the Netjets Assistant Professor at The Ohio State University. She received her B.S. degree in Automation Science from Beihang University and her M.S. and Ph.D. degrees in Aerospace Engineering from Auburn University. After graduation, she worked as an engineer in an automotive technology company, Dynamic Research, Inc., and then joined the University of Washington as a postdoctoral fellow. Dr. Dai’s research focuses on control of autonomous systems, numerical optimization, networked dynamical systems, and space robotics. She is an associate fellow of AIAA and a recipient of the NSF Career Award and NASA Early Faculty Career Award. Dr. Dai is serving as an associate editor of the Journal of Guidance, Navigation, and Control and IEEE Transactions on Aerospace and Electronic Systems.

AEM Colloquium Series: Professor Manuel K. Rausch

The Mechanics of Embedded Fiber Networks: Experiments and Models

Fiber networks embedded in a soft matrix carry load in many materials. They form the backbone of soft tissues such as heart valves and blood clots, and they appear in engineered materials such as fiber-reinforced hydrogels and elastomers, electrospun nonwovens, and soft composites for flexible devices. These materials resemble classical fiber composites, but their fibers are semi-flexible and randomly networked, and their matrix can be orders of magnitude softer. As a result, simple rules of mixtures break down. Yet most of what we know about fiber networks comes from studying them in isolation.

In this talk, I will present our work on how an embedding matrix changes fiber network mechanics. I will begin with elementary studies of individual straight and undulated fibers, modeled as geometrically exact isogeometric beams. Next, I will scale up to free-standing networks. Then I will embed these networks in a hyperelastic matrix using a mortar-type finite element approach. Along the way, I will ask how the matrix changes the way fibers bend, reorient, and stretch, and what that means for stiffness and fracture. I will then turn to experiments, from the mechanics of individual electrospun fibers to composites of these fibers in a soft matrix. We test the composites at the macroscale and use two-photon microscopy to track fiber kinematics with and without the matrix. Finally, I will share preliminary work on a phase-field fracture model that couples the matrix and fiber phases.

Manuel K. Rausch is an Associate Professor of Aerospace Engineering and Engineering Mechanics and of Biomedical Engineering at The University of Texas at Austin, where he directs the Soft Tissue Biomechanics Lab. Originally from Germany, he earned his Ph.D. in Mechanical Engineering from Stanford University in 2013. He then spent two years as Director of R&D at a small medical device company before returning to academia as a postdoctoral fellow at Yale University. He joined UT Austin in 2017. His lab combines experiments and computation to understand how soft tissues such as heart valves, myocardium, and blood clots deform, fail, and remodel. The goal is to improve diagnostics, treatment planning, and medical device design. His honors include an NSF CAREER Award, an American Heart Association Career Development Award, the Moncrief Grand Challenge Award and the Cockrell School’s Award for Outstanding Engineering Teaching.

AEM Colloquium Series: Professor Mihai Duduta

Dielectric Elastomer Actuators for Soft Robots in Extreme Aerospace Environments

Soft machines are built from compliant materials, which gives them the adaptability and resilience of biological organisms when they augment or complement traditional rigid machines. This talk presents the design, fabrication, and flight of a soft robotic system carried to the Stratosphere on a high-altitude balloon. The system is enabled by a UV-curable silicone elastomer mechanism and by dielectric elastomer actuators, which operate as fully solid state compliant capacitors and maintain reliable electromechanical performance from -55 C to 120 C. Gravity, temperature, pressure, and radiation all impact different robotic components in unique ways across different time scales, and the couplings between these axes are largely unmeasured. Reduced gravity is the least understood of these parameters for soft robots, and I will describe how we plan to measure its impact, from materials and interfaces up to complete robots, on a path toward suborbital testing through NASA's RockSat program. The same solid state actuators show promise in other harsh environments, including the deep sea, the Arctic, and nuclear operations.

Mihai "Mishu" Duduta is an assistant professor in the School of Mechanical, Aerospace and Manufacturing Engineering at UConn. He completed a BS in Materials Science and Engineering at MIT, then became the first employee of 24M Technologies, a start-up spun out to commercialize a battery technology he co-invented. His PhD thesis in soft robotics, completed at Harvard University under the guidance of Profs. Robert Wood and David Clarke, included work which won a Gold Award at the Materials Research Society Fall Meeting 2018 and was nominated for Best Paper at ICRA 2018. Before UConn he was a Medical Devices Innovation Fellow at the University of Minnesota, then an assistant professor at the University of Toronto in Mechanical & Industrial Engineering. His research group focuses on materials and manufacturing innovations that enable soft machines to operate in extreme environments. His group recently flew a soft robotic system to the Stratosphere on a high-altitude balloon and is preparing a suborbital demonstration through NASA's RockSat program. Prof. Duduta authored the book Soft Robotics: Building Machines from Soft Matter, published by De Gruyter in 2025.