BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Penn Engineering Events - ECPv6.15.18//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-WR-CALNAME:Penn Engineering Events
X-ORIGINAL-URL:https://seasevents.nmsdev7.com
X-WR-CALDESC:Events for Penn Engineering Events
REFRESH-INTERVAL;VALUE=DURATION:PT1H
X-Robots-Tag:noindex
X-PUBLISHED-TTL:PT1H
BEGIN:VTIMEZONE
TZID:America/New_York
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20230312T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20231105T060000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20240310T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20241103T060000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20250309T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20251102T060000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20240603T130000
DTEND;TZID=America/New_York:20240603T150000
DTSTAMP:20260403T134729
CREATED:20240531T131800Z
LAST-MODIFIED:20240531T131800Z
UID:10007976-1717419600-1717426800@seasevents.nmsdev7.com
SUMMARY:ESE PhD Thesis Defense: "Integrated Electronic-Photonic Solutions From Quantum Control Systems to Optical Transmitters"
DESCRIPTION:Silicon’s advanced fabrication processes have enabled the miniaturization of complex electronic systems\, enhancing performance and efficiency. Recent technological developments have further expanded silicon’s utility to integrate photonic systems\, merging electronic and photonic technologies on a single chip. This integration has opened new avenues for high-speed communication and computation\, attracting significant interest from both research and industry. In this thesis\, integrated electronic-photonic solutions ranging from quantum control systems to optical transmitters are presented. Firstly\, an integrated reconfigurable quantum control system is demonstrated. This system is used to determine electron-spin resonance frequency and perform Rabi\, Ramsey\, and Hahn-echo measurements for an NV center spin qubit in diamond. Secondly\, two monolithically integrated single-channel optical PAM-4 transmitters are implemented\, studied\, and compared. Lastly\, monolithically integrated 8- and 32-channel wavelength-division multiplexed optical transmitter systems are presented. These systems operate in the infrared optical C-band using custom-designed two-section PN-capacitive micro-ring modulators. The 8- and 32-channel systems support aggregate data rates up to 256 Gb/s and 1.024 Tb/s\, respectively\, and are highly integrated with a wavelength stabilization circuit\, test data generators\, and high-swing electrical drivers on the same CMOS silicon-on-insulator chip.
URL:https://seasevents.nmsdev7.com/event/ese-phd-thesis-defense-integrated-electronic-photonic-solutions-from-quantum-control-systems-to-optical-transmitters/
LOCATION:Towne 337
CATEGORIES:Dissertation or Thesis Defense
ORGANIZER;CN="Electrical and Systems Engineering":MAILTO:eseevents@seas.upenn.edu
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20240604T100000
DTEND;TZID=America/New_York:20240604T113000
DTSTAMP:20260403T134729
CREATED:20240507T131428Z
LAST-MODIFIED:20240507T131428Z
UID:10007964-1717495200-1717500600@seasevents.nmsdev7.com
SUMMARY:MEAM Seminar: "Dynamically Tunable Adhesion and Friction via Active Materials with Thermally Modulated Stiffness"
DESCRIPTION:Contact interactions\, including adhesion and friction\, are critical to the design of many engineered systems. Currently\, most systems rely on materials with static mechanical properties\, requiring careful selection of materials to realize effective systems for specialized tasks. However\, with advances in smart materials\, system design is no longer limited to materials with static properties. There is a significant potential to exploit active materials for dynamic control of mechanical behaviors\, including adhesion and friction\, to enable the design of systems with improved performance and new functionalities. Example applications of such systems are robotic grasping and manipulation. In this work\, active control of adhesion and friction is realized using materials with tunable stiffness. In particular\, thermally responsive polymers\, which exhibit substantial changes in stiffness\, provide significant potential for adhesion and friction control. We demonstrate the use of a shape memory polymer with thermally modulated stiffness to dynamically tune adhesion and friction. Through a combination of experimentation and finite element analysis\, we present a composite microstructured adhesive with high strength and adhesion switchability\, while highlighting the role of scale in achieving fast response times. Through further experimentation\, we investigate the ability to tune friction\, using stiffness modulation to enable a transition from Coulomb friction to adhesion-dominated friction. This ability to dynamically control adhesion and friction offers new opportunities for the design of engineered systems.
URL:https://seasevents.nmsdev7.com/event/meam-seminar-dynamically-tunable-adhesion-and-friction-via-active-materials-with-thermally-modulated-stiffness/
LOCATION:Room 337\, Towne Building\, 220 South 33rd Street\, Philadelphia\, PA\, 19104\, United States
CATEGORIES:Seminar,Doctoral
ORGANIZER;CN="Mechanical Engineering and Applied Mechanics":MAILTO:meam@seas.upenn.edu
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20240604T130000
DTEND;TZID=America/New_York:20240604T150000
DTSTAMP:20260403T134729
CREATED:20240528T151953Z
LAST-MODIFIED:20240528T151953Z
UID:10007972-1717506000-1717513200@seasevents.nmsdev7.com
SUMMARY:ESE PhD Thesis Defense: "Algorithms for Adversarially Robust Deep Learning"
DESCRIPTION:Given the widespread use of deep learning models in safety-critical applications\, ensuring that the decisions of such models are robust against adversarial exploitation is of fundamental importance.  In this thesis\, we discuss recent progress toward designing algorithms that exhibit desirable robustness properties.  First\, we discuss the problem of adversarial examples in computer vision\, for which we introduce new technical results\, training paradigms\, and certification algorithms.  Next\, we consider the problem of domain generalization\, wherein the task is to train neural networks to generalize from a family of training distributions to unseen test distributions.  We present new algorithms that achieve state-of-the-art generalization in medical imaging\, molecular identification\, and image classification.  Finally\, we study the setting of jailbreaking large language models (LLMs)\, wherein an adversarial user attempts to design prompts that elicit objectionable content from an LLM.  We propose new attacks and defenses\, which represent the frontier of progress toward designing robust language-based agents.
URL:https://seasevents.nmsdev7.com/event/ese-phd-thesis-defense-algorithms-for-adversarially-robust-deep-learning/
LOCATION:Wu and Chen Auditorium (Room 101)\, Levine Hall\, 3330 Walnut Street\, Philadelphia\, PA\, 19104\, United States
CATEGORIES:Dissertation or Thesis Defense
ORGANIZER;CN="Electrical and Systems Engineering":MAILTO:eseevents@seas.upenn.edu
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20240604T140000
DTEND;TZID=America/New_York:20240604T150000
DTSTAMP:20260403T134729
CREATED:20240516T141242Z
LAST-MODIFIED:20240516T141242Z
UID:10007969-1717509600-1717513200@seasevents.nmsdev7.com
SUMMARY:GRASP Seminar: Lillian Chin\, University of Texas at Austin\, "Materials Make the Bot: Directly Embedding Actuation and Perception into Robotic Structures"
DESCRIPTION:*This seminar will be held in-person in Levine 307 as well as virtually via Zoom.  \nABSTRACT\n\nTo make a future where robots are helpful and commonplace\, robots must physically interact with humans and their surroundings. In this talk\, I argue that robots should be designed from a materials-centric approach to better facilitate these interactions. If core robotic features like actuation and perception can be directly incorporated into a robot’s materials\, we could directly control the robot’s primary interface to the outside world. \nDrawing from principles in mathematics and metamaterial design\, I use auxetic materials as a case study to show how metamaterials can be explicitly designed as the foundation for a robot’s movement and sensing capabilities. I demonstrate the power of this approach by creating expanding modular robots with strength-to-weight ratios of 76x and developing a novel class of auxetics that make soft robotic grippers that are 20x more efficient than standard pneumatic versions. I also present a method for directly sensorizing metamaterial structures in general by embedding internal fluidic channels within the struts themselves as the structure is being 3D printed. This technique offers proprioceptive feedback with minimal hysteresis\, enabling accurate pose reconstruction from these fluidic sensors alone. I close my talk with some preliminary work on adapting this materials-focused approach towards medical applications.
URL:https://seasevents.nmsdev7.com/event/grasp-seminar-lillian-chin-university-of-texas-at-austin-materials-make-the-bot-directly-embedding-actuation-and-perception-into-robotic-structures/
LOCATION:Levine 307\, 3330 Walnut Street\, Philadelphia\, PA\, 19104\, United States
CATEGORIES:Seminar
ORGANIZER;CN="General Robotics%2C Automation%2C Sensing and Perception (GRASP) Lab":MAILTO:grasplab@seas.upenn.edu
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20240605T140000
DTEND;TZID=America/New_York:20240605T153000
DTSTAMP:20260403T134729
CREATED:20240529T174333Z
LAST-MODIFIED:20240529T174333Z
UID:10007973-1717596000-1717601400@seasevents.nmsdev7.com
SUMMARY:xLab Seminar: "Learning to Control with Vision–Language Models"
DESCRIPTION:If learning from data is valuable\, can learning from big data be very valuable? It has been\, so far\, in vision and language\, for which foundation models can be trained on web-scale data to support a plethora of downstream tasks; not so much in control\, for which scalable learning remains elusive. Can information encoded in vision and language models guide reinforcement learning of control policies? In this talk\, I will discuss several ways for foundation models to help agents to learn to behave. Language models can provide better context for decision-making: we will see how they can succinctly describe the world state to focus the agent on relevant features; and how they can form generalizable skills that identify key subgoals. Vision and vision–language models can help the agent to model the world: we will see how they can block visual distractions to keep state representations task-relevant; and how they can hypothesize about abstract world models that guide exploration and planning.
URL:https://seasevents.nmsdev7.com/event/xlab-seminar-learning-to-control-with-vision-language-models/
LOCATION:Towne 337
CATEGORIES:Seminar,Colloquium
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20240605T153000
DTEND;TZID=America/New_York:20240605T163000
DTSTAMP:20260403T134729
CREATED:20240531T150349Z
LAST-MODIFIED:20240531T150349Z
UID:10007977-1717601400-1717605000@seasevents.nmsdev7.com
SUMMARY:IDEAS Seminar: "An optimization framework for designing robust state estimators"
DESCRIPTION:Cyber-physical systems often include communication networks that ensure data transmission between different components of the system (sensors\, actuators\, processing units\, etc). The presence of such networks renders the whole system vulnerable to malicious attacks consisting\, for example\, in the injection of arbitrary signals. In this context\, the data collected over the communication channel may be so unreliable that their use for state estimation or system identification requires design methods which are more robust than conventional ones. \nResilience is a particular robustness property which characterizes the sensitivity of some performance function of interest with respect to a class of disturbances (model uncertainties). For example\, we say that a state estimator is resilient to a set of disturbances E if the estimation error induced by that estimator is (a) zero whenever the actual model uncertainty lies in E and (b) continuously dependent on the distance from the actual uncertainty to the set E. In this talk we will discuss a resilience-inducing optimization framework for secure state estimation in the scenario where E is a set of impulsive (or sparse) noise sequences. This type of noise signal can account typically for intermittent sensor failures or adversarial attacks in the context of cyber-physical systems. It can also arise artificially as a methodological device for example\, in the identification\, estimation and control of switched systems. We consider both batch off-line and online recursive estimation.
URL:https://seasevents.nmsdev7.com/event/ideas-seminar-an-optimization-framework-for-designing-robust-state-estimators/
LOCATION:Levine 307\, 3330 Walnut Street\, Philadelphia\, PA\, 19104\, United States
CATEGORIES:Seminar,Colloquium
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20240607T113000
DTEND;TZID=America/New_York:20240607T133000
DTSTAMP:20260403T134729
CREATED:20240523T145908Z
LAST-MODIFIED:20240523T145908Z
UID:10007970-1717759800-1717767000@seasevents.nmsdev7.com
SUMMARY:BE Doctoral Dissertation Defense: "Goal-Directed Dynamics of Network Topology" (Shubhankar Patankar)
DESCRIPTION:The Department of Bioengineering at the University of Pennsylvania and Dr. Dani Smith Bassett are pleased to announce the Doctoral Dissertation Defense of Shubhankar Patankar.\n\n\nTitle: Goal-Directed Dynamics of Network Topology\nDate: June 7\, 2024\nTime: 11:30 am – 1:30 pm\nLocation: Heilmeier Hall\, Room 100 Towne Building\nZoom: https://upenn.zoom.us/j/95495692170\n\nThe public is welcome to attend.
URL:https://seasevents.nmsdev7.com/event/be-doctoral-dissertation-defense-goal-directed-dynamics-of-network-topology-shubhankar-patankar/
LOCATION:Heilmeier Hall (Room 100)\, Towne Building\, 220 South 33rd Street\, Philadelphia\, PA\, 19104\, United States
CATEGORIES:Doctoral,Graduate,Student,Dissertation or Thesis Defense
ORGANIZER;CN="Bioengineering":MAILTO:be@seas.upenn.edu
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20240607T130000
DTEND;TZID=America/New_York:20240607T130000
DTSTAMP:20260403T134729
CREATED:20240605T192816Z
LAST-MODIFIED:20240605T192816Z
UID:10007983-1717765200-1717765200@seasevents.nmsdev7.com
SUMMARY:infoLeNS Seminar: "Emerging computational imaging inverse problems: from theory to algorithms"
DESCRIPTION:In this talk\, I will focus on two challenging imaging systems: snapshot compressive imaging and coherent imaging under speckle noise interference. I will begin by reviewing the core mathematical modeling of the inverse problem corresponding to each system. I will develop a maximum likelihood estimator (MLE)-based optimization for each\, employing untrained neural networks (NNs) to model the source structure. Theoretical analysis of the MLE-based methods will be shown to enable\, on one hand\, an understanding of the fundamental limits of these systems and\, on the other hand\, optimization of the image recovery algorithms and hardware. I will also discuss our proposed algorithms that merge classic bagging ideas with untrained neural networks for solving the inverse problems in these imaging systems. For each application\, I will demonstrate how our method achieves state-of-the-art performance.
URL:https://seasevents.nmsdev7.com/event/infolens-seminar-emerging-computational-imaging-inverse-problems-from-theory-to-algorithms/
LOCATION:Room 452 C\, 3401 Walnut\, 3401 Walnut Street\, Philadelphia\, PA\, 19104\, United States
CATEGORIES:Seminar,Colloquium
END:VEVENT
END:VCALENDAR