The reproductive soft tissues that support the fetus undergo some of the most dramatic and unique growth and remodeling events in the human body. The uterus and fetal membrane must grow and stretch during pregnancy to accommodate the fetus. Simultaneously, the cervix must remodel and be a mechanical barrier to keep the fetus within the […]
For the past two decades, power management and delivery has been a bottleneck, limiting the size and performance of a range of applications from performance computing to mobile phones and wearables. In the coming decades, power electronics will become the ‘glue’ of the modern energy system network. With electronics embedded deeply into this network, well […]
System software like operating systems and hypervisors forms the critical backbone of our computing infrastructure. However, due to their size and complexity, these systems often contain vulnerabilities that can compromise security. Formal verification offers a solution by mathematically proving software correctness, but its adoption is hindered by the substantial effort required to create these proofs. […]
Abstract: Transformer-based language models have undoubtedly become the dominant and favorite architecture for language generation of our time. However, although they provide impressive text quality, they tend to be hard to control. In the domain of image synthesis, on the other hand, Denoising Diffusion Models (DDM) are the dominant approach, shining with unprecedented quality and […]
This will be a hybrid event with in-person attendance in Levine 307 and virtual attendance on Zoom. ABSTRACT Safe and efficient motion planning is crucial for robots operating in dynamic and unstructured environments. To accomplish this, one must address key challenges. First, robots must understand the scene geometry to prevent collisions that could harm humans […]
Connecting theoretical models for exotic quantum states to real materials is a key goal in quantum materials science. The structure of the crystalline lattice plays a foundational role in this pursuit in the subfield of quantum material synthesis. We here revisit this long-standing perspective in the context low dimensional emergent electronic phases of matter, including […]
This will be a hybrid event with in-person attendance in Wu and Chen and virtual attendance on Zoom. ABSTRACT Atomic systems (molecules, crystals, proteins, etc.) are naturally represented by a set of coordinates in 3D space labeled by atom type. This poses a challenge for machine learning due to the sensitivity of coordinates to 3D […]
Digital twins are virtual models of physical systems that allow for more computationally cost-effective evaluation and optimization. Building digital twins often involves machine learning techniques that integrate data with underlying physical laws. In this seminar, I'll explore two such techniques: Physics-Informed Neural Networks (PINNs) and operator learning. First, I'll discuss the formulation of PINNs and […]
The Department of Bioengineering at the University of Pennsylvania and Dr. Peter Noel are pleased to announce the Doctoral Dissertation Defense of Leening Liu. Title: Spectral CT thermometry for real-time, non-invasive temperature monitoring of thermal ablation Date: November 4, 2024 Time: 2:00 PM Location: Reunion Auditorium, John Morgan Building Zoom info: Topic: Leening Liu's Dissertation Defense Time: November 4, […]
Funicular structural forms maximize the structural performance and minimize the use of materials by carrying the applied loads in the form of pure tensile/compressive axial forces. The internal structure of a bone is a classic example where material follows the principal stress directions and forms a delicate latticework of tiny, interlaced trabeculae crossing each other. […]
I would probably "forget" what I will say in the abstract, and you would, too. Translation: deep neural networks can "forget", meaning they might perform poorly on previously learned tasks when learning a new task. A major goal of the subject now known as deep continual learning is to address this issue. In order to […]
Abstract: Modern machine learning methods are revolutionizing what we can do with data, from tiktok video recommendations to biomarkers discovery in cancer research. Yet, the complexity of these deep models makes it harder to understand what functions these data-dependent models are computing, and which features they detect regarding as important for a given task. In this talk, […]
Abstract: The growing need to secure critical minerals for clean energy technologies, alongside the urgency to mitigate climate change, presents a unique opportunity to combine mineral recovery with carbon sequestration. This dissertation explores the potential of mine tailings as a dual-purpose feedstock, enabling both the extraction of critical minerals and the storage of CO₂ through […]
This will be a hybrid event with in-person attendance in Levine 307 and virtual attendance on Zoom. ABSTRACT Recent work has demonstrated that a promising strategy for teaching robots a wide range of complex skills is by training them on a curriculum of progressively more challenging environments. However, developing an effective curriculum of environment distributions […]
There are many situations in which small particles impact metals at high speeds, even above the speed of sound. Sometimes these are unintentional (as for foreign object damage or micrometeorite strikes), and sometimes they occur by design (as for surface treatment processes like abrasive spray, peening, or spray coatings). The fundamental physics behind supersonic impacts, […]
The BU Wireless Integrated Systems and Extreme Circuits (WISE-Circuits) group blends integrated circuits with energy-constrained applications to pioneer breakthroughs in information theory, bioengineering, and communications. Our research focuses on the development of Cyber-Secure Biological Systems, leveraging living sensors constructed from engineered biological entities seamlessly integrated with solid-state circuits. This unique synergy harnesses the advantages of […]
Cedar is a new open-source authorization policy language, used to express fine-grained permissions on behalf of applications. Rather than embed authorization logic in their application code, developers can write that logic as Cedar policies and delegate access requests to Cedar’s evaluation engine. Cedar is designed to be ergonomic, fast, safe, and analyzable. Cedar’s simple and […]
This will be a hybrid event with in-person attendance in Wu and Chen and virtual attendance on Zoom. ABSTRACT In the last few years, the ability for robots to understand and operate in the world around them has advanced considerably. Examples include the growing number of self-driving car systems, the considerable work in robot mapping, […]
Despite substantial progress in robotics, achieving human-like manipulation remains a significant challenge. Existing robotic systems typically leverage human-inspired sensory modalities: vision, touch, and proprioception. However, these modalities are historically studied and integrated in isolation, leading to limited performance in complex real-world tasks that require sensing across multiple modalities for robust generalization. As a result, robots […]
Amorphous materials have no long-range order, but there are ordered structures at short-range (2-5 Å), medium-range (5-20 Å), and even longer-length scales. While regular and semiregular polyhedra are often identified as short-range order in amorphous materials, the nature of the medium-range order has remained elusive. Because of the disorder, the dynamics also become far more […]
Mobile robots have shown significant advancements in agility, intelligence, and efficiency over the past few decades. However, their endurance and overall performance remain limited by the onboard power supplies. Current power sources typically restrict mobile robots to areas close to the electrical grid and necessitate heavier batteries for extended range. Energy refueling could be significantly […]
When a superhydrophobic (SH) textured surface is immersed in water it traps a thin shiny layer of air within the texture that is known as a plastron. Contact line pinning stabilizes this Cassie-Baxter state and the patches of air trapped in the texture can act as shear-free regions that locally lower the frictional dissipation. Recent […]
Color center systems are among the leading platforms in the development of quantum communication and quantum sensing hardware due to their desirable spin, optical, and spin-photon properties. Among them, the near infrared emitters in silicon carbide, such as the nitrogen-vacancy center in 4H-SiC, provide fiber-friendly operation in an industrially mature substrate, ideal for scalable deployment […]
Mechanical metamaterials are artificially designed structures that exhibit unique properties due to their internal structure rather than their composition, e.g., negative Poisson’s ratio, tunable stiffness, and advanced thermal characteristics. While the static properties of mechanical metamaterials have been widely studied, their nonlinear dynamics remain largely unexplored, which could pave ways for innovative design and optimization […]
As Artificial Intelligence (AI) finds increasing applications in industry and society. Responsible deployment demands that we measure and correct algorithmic biases vis-a-vis protected attributes such as sex, age and ethnicity. State of the art methods for measuring algorithmic bias rely on test sets that are collected in the wild and are then annotated for the […]
Abstract: Tumors are complex and heterogeneous systems, which challenge their classification and treatment. The Silverbush lab decodes tumor heterogeneity and plasticity to understand how cancer cells transform to become more aggressive or evade treatment. We particularly examine hard-to-treat cancers with high heterogeneity, focusing on the notorious kings of heterogeneity and aggressiveness: brain cancers. To achieve this […]
The theory of Coherent Perfect Absorption1 (time-reversed lasing) and Reflectionless Scattering Modes2 has shown that there always exist discrete solutions to scattering of electromagnetic waves in multiple-scattering geometries that achieve perfect transduction or perfect (reflectionless) impedance matching. In general these are transient (complex frequency) solutions, but, typically, with tuning of a one system parameter it […]
A crucial yet unavailable component in high-performance photonic integrated circuits (ICs) and other chip-scale photonic systems is an on-chip light source that is efficient, functional, IC-compatible, and electronically addressable. In this talk, I will cover several types of on-chip sources, including perovskite microlasers and luminescent hyperbolic metamaterials, topologically protected microlasers on the III-V platform, as […]
*This seminar will be held in-person in Levine 307 as well as virtually via Zoom. This seminar will NOT be recorded. ABSTRACT This talk has two parts. First, I will explain our recent work on making it easier to control 3D soft robots. By extending our prior work on soft robot inverse kinematics (Soft IK) into […]
This will be a hybrid event with in-person attendance in Wu and Chen and virtual attendance on Zoom. ABSTRACT In 2022 I represented Agility Robotics at the first “Philadelphia Summit” on the weaponization of autonomous robots, co-hosted by the GRASP lab and Boston Dynamics. By late fall, this yielded a joint statement, signed by Agility, […]
Synchronized operation of a large number of similar entities can generate behavior that cannot be expected from single one. This principle combined with the physical concept of coherence in waves enables creation of programmable 3-dimensional focal points that can be used to enable many applications such as wireless energy transfer at distance on earth and […]
*This seminar will be held in-person in Levine 307 as well as virtually via Zoom. ABSTRACT Foundation models, such as GPT-4 Vision, have marked significant achievements in the fields of natural language and vision, demonstrating exceptional abilities to adapt to new tasks and scenarios. However, physical interaction—such as cooking, cleaning, or caregiving—remains a frontier where foundation […]
Learning-based model predictive control (MPC) is an increasingly prominent control paradigm in recent years. One primary approach in learning-based MPC is to leverage machine or deep learning tools to construct accurate dynamics models. These models are then deployed into optimization-based control schemes. Although these frameworks can potentially provide significant performance improvements, a number of key […]
Fluorinated polymers offer a plethora of unique and sometimes perplexing properties. One of the most interesting is the ability for specific types of polytetrafluoroethylene (PTFE) to undergo room-temperature, sheer-induced, structural transformations leading “unwinding” of crystallites and the formation of nano-scale fibers. This process is commonly referred to as “fibrillation”. The material properties required for this […]
This talk presents novel non-parametric methods for analyzing dynamical systems using solely trajectory data. Our critical insight is to replace the notion of invariance, a core concept in Lyapunov Theory, with the more relaxed condition of recurrence. Specifically, a set is τ-recurrent if every trajectory that starts within the set returns to it after at […]
Sociotechnical systems enable large-scale connection and the dissemination of timely, engaging information. However, these systems are shaped by complex, often invisible governance structures—comprising rules, roles, and relationships—that significantly influence how they operate. Gaining even a basic understanding of these governance structures is key to using them more effectively and deliberately. Yet, many of the millions […]
Abstract: Performance of control systems depend on the nature of available measurements. Perception of edges, keypoints, landmarks and other natural semantic features make certain coarse measurements available to control systems operating in complex environments. This talk explores two problems related to control and estimation with such coarse measurements. First, I will introduce perception contracts—an approach for analyzing visual […]
In this talk I will tell you how scaling has been the engine of progress in AI for the past five years. In the first scaling paradigm, our field scaled large language models by training with more compute on more data. Such scaling led to the success of ChatGPT and other AI chat engines, which were surprisingly capable […]
This is an in-person event only and is not open to the public. Event Registration and Penncards are required. A link to register for the event will be sent via email. PANEL DISCUSSION Please join us for a panel discussion featuring the 2024 John Scott Award Recipients, Dr. Takeo Kanade (CMU), Dr. Vijay Kumar (Penn), […]
Machinery relies on lubrication to regulate friction and wear at contacting interfaces. As lubricants become less viscous to save energy and cost, and as new technologies like electric vehicles operate in harsher conditions, the risk of surface-initiated failure grows. We show that metal oxide nanocrystals (NCs) dispersed in lubricants form protective coatings, or tribofilms, in […]
Mineralized tissues are paradigmatic hierarchical materials that reap synergy from structural and compositional gradients at multiple length scales in ways that are challenging to reproduce by conventional means. My laboratory studies the formation, functional properties, and degradation of mineralized tissues. We use model systems ranging from single crystalline endoskeletal elements deposited by single cells to […]
The electric grid is undergoing a transformative paradigm shift, driven by sweeping changes in generation, demand, and energy storage. By 2035, solar PV alone is expected to supply 40% of U.S. electricity, with substantial additional contributions from wind, geothermal, and hydroelectric sources—creating a renewable-dominant energy landscape. Meanwhile, electricity demand is accelerating due to rapid growth […]
This will be a hybrid event with in-person attendance in Wu and Chen and virtual attendance on Zoom. ABSTRACT Nature’s musculoskeletal design can inspire both artificial and living robots to create systems that can better interact within our unstructured world. There is value in rethinking how we design and control robots by replacing traditional designs […]
The intersection of machine learning (ML) and dynamical systems and control (S&C) has become a prominent area of research in recent years. While most work applies ML to S&C problems, this dissertation explores the reverse direction: using S&C tools to address challenges in ML and optimization. This thesis is comprised of three parts: In part […]
Research in the Zahrt group focuses on creating new tools to advance organic synthesis by integrating automation and machine learning workflows to enhance molecular function, reaction efficiency, and sustainability. We develop and use active learning strategies for catalyst design, reaction conditions, and other molecular properties. To accelerate the implementation of these algorithms, we pair them […]
To improve FPGA design productivity, our goal is to create a development experience for FPGAs that aligns closely with widely accepted software design principles. Software programmers quickly test their minimally completed design, identify the bottleneck, and incrementally refine the design. In FPGA design, however, such incremental refinement is not currently supported. (1) FPGA compilation is long, […]
