Colloquium
CS Colloquiums and The UNM Computer Science Cleve Moler & MathWorks Chair in Mathematical and Engineering Software Distinguished Lecture Series talks are held Wednesdays from 2-3:00 p.m. in the CEC 1041 Auditorium.
2026 Computer Science Colloquium Series
Interplay Between Learning and Control: Robustness of Learning Algorithms and Data-Driven Controller Design
Leilei Cui, University of New Mexico
Wednesday, April 1, 2026, 2:00 PM
Location: Larrañaga Engineering Auditorium (Centennial 1041)
Abstract:
Learning and control are deeply intertwined disciplines. On one hand, control theory provides a rigorous mathematical framework for understanding, analyzing, and certifying learning algorithms. On the other hand, learning enables data-driven, model-free controller design for dynamical systems.
At the core of most learning methods lie gradient-based optimization algorithms, whose performance in practice is inevitably affected by disturbances such as data noise and computational inaccuracies. These perturbations raise a fundamental question: Can optimization algorithms maintain convergence to near-optimal solutions in the presence of disturbances? To address this question, we leverage tools from control theory—particularly input-to-state stability (ISS)—to rigorously analyze the robustness of gradient-based optimization algorithms under disturbances. We show that, under mild assumptions, these algorithms converge to a small neighborhood of the optimal solution, provided that the perturbations remain bounded and sufficiently small.
In addition, we highlight the power of learning in controller design, especially for time-delay systems, where computing optimal controllers is challenging even when accurate models are available. In such settings, we develop a direct data-driven approach that learns optimal controllers directly from data. Finally, we demonstrate the effectiveness of learning-based control through applications in robotics.
Bio:
Leilei Cui is an Assistant Professor in the Department of Mechanical Engineering at the University of New Mexico. He was a Postdoctoral Associate at the Massachusetts Institute of Technology (MIT) from June 2024 to July 2025. He received his M.Sc. degree in Control Science and Engineering from Shanghai Jiao Tong University, China, in 2019, and his Ph.D. degree in Electrical Engineering from New York University in 2024.
Dr. Cui’s research lies at the intersection of control theory, optimization, and reinforcement learning, with a particular focus on applications to robotic systems. He is the recipient of the Outstanding Paper Award from the journal Control Theory and Technology, the Dante Youla Award for graduate research excellence, and the Alexander Hessel Award for the best Ph.D. dissertation in Electrical Engineering at NYU.
Data-Driven Discovery and Prediction of Wildfire Impacts on Fluvial Networks
Ricardo Gonzalez-Pinzon, University of New Mexico
Wednesday, March 11, 2026, 2:00 PM
Location: Larrañaga Engineering Auditorium (Centennial 1041)
Abstract:
Wildfires are increasingly generating large-scale hydrologic and biogeochemical disturbances that propagate through river networks, altering runoff regimes, sediment transport, and nutrient export over hundreds of kilometers. Despite expanding sensor deployments and public datasets, scalable integration frameworks and generalizable predictive models for post-fire impacts remain limited. In this seminar, I present three computational products that address this challenge: RIO-FINDER, TIERRAS, and a multilayer perceptron (MLP) surrogate model for solute transport. RIO-FINDER (River Integrated Observations for Fire-Impacted Network Discovery and Evaluation in R) is an open-source geospatial framework that links four decades of wildfire perimeters (MTBS) with high-resolution hydrologic observations (USGS NWIS) through the NHDPlus stream topology. It performs large-scale spatial intersection analyses to identify monitoring sites embedded within fire-affected basins, enforce burn-percentage thresholds, and filter for sites with both pre- and post-fire observations. The result is a CONUS-scale inventory accessible through an interactive Shiny interface, along with a rapid-analysis tool for newly burned watersheds. Complementing this discovery framework, TIERRAS (Tracer Injection Experiments in Rivers and Streams) is a curated database of conservative solute transport experiments spanning diverse hydrogeomorphic settings. The dataset consolidates tracer breakthrough curves (BTCs), standardizes heterogeneous metadata, and augments observations with physically consistent synthetic solutions of the advection–dispersion equation. Building on TIERRAS, we develop a pretrained MLP model that predicts breakthrough curves without site-specific calibration, leveraging synthetic pretraining and fine-tuning on experimental data to achieve robust cross-site generalization. Together, these tools form an end-to-end pipeline for scalable discovery, data fusion, and machine-learned forecasting of wildfire-driven disturbances across continental-scale fluvial networks.
Bio:
Ricardo González-Pinzón, Ph.D., is a Professor of Water Resources Engineering at the University of New Mexico and Interim Director of the Center for Water and the Environment. His research integrates hydrology, environmental sensing, data analytics, and predictive modeling to understand how disturbances (e.g., wildfires, floods, and contamination events) propagate through river networks. Ricardo’s team aims to develop scalable, open computational tools and machine learning frameworks to detect, quantify, and predict impacts on water quantity, quality, and ecosystem processes. They also lead rapid-response research efforts following wildfires and advance high-resolution monitoring technologies to improve environmental observability.
AI for Health: Translating Data into Impact
Chenyang Lu, Washington University in Saint Louis
Wednesday, March 4, 2026, 2:00 PM
Location: Larrañaga Engineering Auditorium (Centennial 1041)
Abstract:
Health care is one of the most exciting, complex, and high-stakes domains for AI. This talk highlights recent advances in AI for health, focusing on how progress in machine learning, foundation models, and multimodal learning can be translated into real-world impact. Through examples spanning clinical care and public health, the talk examines the technical challenges and solutions that arise when AI meets messy, diverse, real-world data. It concludes by outlining interdisciplinary research opportunities where advanced AI can meaningfully enhance human health.
Bio:
Chenyang Lu is the Fullgraf Professor of Computer Science & Engineering at Washington University in St. Louis, with joint appointments in Anesthesiology, Medicine, Neurosurgery, and Public Health. He is the founding director of the AI for Health Institute, where he leads a multidisciplinary initiative that brings together AI researchers and health professionals to address critical challenges in health care and public health through data-driven innovation. His research leverages advanced AI and machine learning approaches to advance both precision medicine and population health. Dr. Lu is a Fellow of the ACM and IEEE, received the 2022 Outstanding Technical Achievement and Leadership Award from the IEEE Technical Community on Real-Time Systems, and serves as Editor-in-Chief of ACM Transactions on Cyber-Physical Systems.
Strategies to Suppress and Correct Noise in Quantum Systems
Milad Marvian, University of New Mexico
Wednesday, February 25, 2026, 2:00 PM
Location: Larrañaga Engineering Auditorium (Centennial 1041)
Abstract:
The susceptibility of quantum systems to noise remains a major obstacle to realizing their full potential for computational tasks. While the threshold theorem guarantees that arbitrarily long quantum computations can be performed reliably provided each component's noise level remains below a certain threshold, meeting the overhead and noise-level requirements poses significant technological challenges. In this talk, I will review some recently developed strategies to suppress and correct quantum noise. I will discuss easy-to-implement techniques such as randomized dynamical decoupling and energy-penalty error suppression to reduce the noise on each qubit, as well as design principles for low-overhead fault-tolerant quantum circuits.
Bio:
Milad Marvian an Assistant Professor in the Department of Electrical & Computer Engineering at the University of New Mexico and also a member of the Center for Quantum Information and Control (CQuIC). His research interests include the theoretical aspects of quantum computation, with a focus on quantum error correction and algorithms. He is also interested in exploring the connections to quantum optimal control and open quantum systems. Prior to joining UNM in 2020, he was a postdoctoral associate at MIT, and before that, he completed his Ph.D. in Electrical & Computer Engineering at the University of Southern California in 2018.
Exploring Unanticipated Functionality in the Mobile Ecosystem and Beyond
Kevin Butler, University of Florida
Wednesday, February 18, 2026, 2:00 PM
Location: Larrañaga Engineering Auditorium (Centennial 1041)
Abstract:
Telecommunication networks form the backbone of our connected society, enabling global voice and data communication. Yet, beneath their seamless operation lies a complex interplay of signaling protocols, legacy systems, and evolving technologies that, over the decades, have exposed both opportunities for innovation and significant security challenges. While these networks have changed over the past 50 years and complexity has migrated outwards to devices, one enduring challenge has been a lack of accessibility. This talk will cover our efforts to better assess security in the mobile device ecosystems, from platforms to core cellular infrastructure. The theme of unanticipated functionality - and the consequences of such functionality on users - is common in security. I will discuss this theme in the context of marginalized and vulnerable users, whose security needs have been largely unaddressed, and offer some thoughts on emerging security challenges for computer systems and users.
Bio:
Kevin Butler is the Director of the Florida Institute for Cybersecurity Research and a UF Research Foundation Professor in the Department of Computer and Information Science and Engineering at the University of Florida. Kevin's research focuses on the security of computer systems as well as the security and privacy of users. He is the PI of a National Science Foundation Frontiers center-scale project on privacy and security for marginalized and vulnerable populations (PRISM). He received an NSF CAREER award in 2013 and was co-chair of the International Telecommunication Union's Security, Infrastructure, and Trust Working Group within the Financial Inclusion Global Initiative from 2016-2022. His work has received multiple awards including from ACM CCS, USENIX Security, ACM CODASPY, and other venues. Kevin is an Executive Committee member of the CRA's Community Computing Consortium and was named to a PCAST review working group.
Trustworthy AI and Autonomy: A Formal Methods–Driven Approach
Pavithra Prabhakar, University of New Mexico
Wednesday, February 11, 2026, 2:00 PM
Location: Larrañaga Engineering Auditorium (Centennial 1041)
Abstract:
Autonomous and cyber-physical systems (CPS) integrate software tightly with physical processes and are increasingly deployed across a wide range of application domains. Modern autonomous CPS depend critically on AI-enabled components for perception, control, and decision making, especially machine-learning and neural-network–based methods. In safety-critical settings—such as aerospace, robotics, and autonomous transportation—failures of these components can have catastrophic consequences, making a high level of trustworthiness essential.
Our lab conducts research on the rigorous analysis and assurance of AI-enabled autonomous systems using formal methods, a class of mathematically rigorous techniques. We advocate abstraction as a foundational principle for addressing the inherent scalability challenges posed by high-dimensional AI models and their interaction with continuous physical dynamics. By constructing sound, yet computationally tractable abstractions, we enable formal verification and validation of system-level safety, stability and correctness properties.
We demonstrate the effectiveness of this approach through multiple real-world case studies, including the verification of neural networks used in collision-avoidance protocols and autonomous rocket-landing systems, among others. These examples illustrate how formal methods can deliver strong, mathematically grounded guarantees for AI-based autonomy, advancing the development of trustworthy, reliable, and deployable autonomous CPS.
Bio:
Pavithra Prabhakar is a Professor in the Department of Computer Science and the Cleve Moler and MathWorks Endowed Chair in Mathematical and Engineering Software at the University of New Mexico. Prior to joining UNM, she was a Professor of Computer Science and the Peggy and Gary Edwards Chair in Engineering at Kansas State University. She earned her Ph.D. in Computer Science and an M.S. in Applied Mathematics from the University of Illinois Urbana–Champaign, followed by a CMI Postdoctoral Fellowship at the California Institute of Technology.
Prabhakar’s research focuses on formal methods for AI-enabled autonomous, cyber-physical, and robotic systems, with applications in aerospace, automotive, and agricultural automation. She has authored over 100 peer-reviewed publications and has received numerous honors, including a Marie Curie Career Integration Grant from the EU, NSF CAREER Award, ONR Young Investigator Award, NITW Distinguished Young Alumnus Award, Amazon Research Award, CRA Future Leader recognition, and a 2025 Early Career Academic Achievement Alumni Award from UIUC.
More recently, she served as a Program Director at the National Science Foundation (NSF) in the CISE Directorate, where she led a portfolio of over 200 research projects with a total budget exceeding $100 million, spanning Formal Methods, Cyber-Physical Systems, Robotics, and Artificial Intelligence.
Heterochiral DNA for intracellular Computing
Matthew Lakin, University of New Mexico
Wednesday, February 4, 2026, 2:00 PM
Location: Larrañaga Engineering Auditorium (Centennial 1041)
Abstract:
Many societal challenges, including the diagnosis and treatment of disease, can be tackled by harnessing the unique capabilities of biology. A key goal in biomedical engineering is to improve human health via personalized biomedicine, which promises targeted treatment guided by the patient's physiology. To this end, I work to engineer and build programmable biological systems for autonomous sensing, decision-making, and response in living organisms. I adopt an interdisciplinary approach, combining experimental validation with formal computational methods for the specification, compilation, and verification of designs. In this talk, I will outline some of my contributions, focusing on domain-specific languages for computational biodesign and on the use of heterochiral DNA, which combines both naturally occurring D-DNA and chiral mirror image L-DNA, to build robust experimental nanodevices for sensing and decision-making in the harsh environment of a living cell.
Bio:
Matthew Lakin obtained his B.A. and Ph.D. in Computer Science from the University of Cambridge. After graduating, he worked as a Postdoctoral Researcher in the Biological Computation Group at Microsoft Research in Cambridge where he developed Visual DSD, a popular software tool for DNA circuit design. Dr. Lakin is currently an Associate Professor in the Department of Computer Science at the University of New Mexico, where he also holds a courtesy appointment in the Department of Chemical & Biological Engineering. His interdisciplinary group works on computational and experimental aspects of DNA nanotechnology, molecular computing, and synthetic biology. Dr. Lakin received the NSF CAREER award in 2021 and the Presidential Early Career Award for Scientists and Engineers (PECASE) in 2025. He has also previously won the UNM School of Engineering's Junior Faculty Research Excellence Award.
Enabling Human-centered Machine Automation with Bi-directional Teaming Integration of Dynamics using Immersive Digital Twins
Fernando Moreu, University of New Mexico
Wednesday, January 28, 2026, 2:00 PM
Location: Larrañaga Engineering Auditorium (Centennial 1041)
Abstract:
This seminar challenges the term automation in the context of engineering experiments and vibrations, proposing new human-computer interfaces and human-in-the-loop control paradigms. Applications include using Augmented Reality (AR) systems enabling a standalone human interface with Digital Twins (DT) for automatic defect detection and non-linear characterization of vibratory systems. Applications include industrial human-centered machine-enabled inspection with applications for smart manufacturing DT. Applications of this new machine-aided DT inspection employs a nonstationary pixel unit conversion with an automated image conversion translating anomalies to engineering units in the eyes of the inspector. The proposed interface with DT also enables intuitive robot programming adjusting the kinematic controller’s parameters with application for construction, manufacturing, and field inspections. The practical application and validation of the model is examined through experiments involving human upper limb movement in interaction with robots. This immersive AR interface opens a bi-lateral communication line between humans and robots for collaboration and bi-directional DT teaming. Ongoing practical applications include the use of neuromorphic, low-latency system identification for non-linear systems ID, and human-robot construction teaming with the help of AR DT.
Bio:
Fernando Moreu, PE is the Robert J. Stamm Professor of Advanced Design and Construction Practices, Dean’s Excellence Lecturer and Associate Professor at the Department of Civil, Construction, and Environmental Engineering (CCEE) at the University of New Mexico (UNM). He holds courtesy appointments in the Departments of Electrical & Computer Engineering, Mechanical Engineering, and Computer Science at UNM and the founder and director of the Smart Management of Infrastructure Laboratory (SMILab, https://smilab.space/). Prof. Moreu received his MS and PhD degrees in structural engineering from the University of Illinois at Urbana-Champaign (2005 and 2015, respectively). Prof. Moreu’s research interests include structural dynamics and control, structural health monitoring, wireless smart sensor networks, cyber-physical systems, computer vision, augmented reality, unmanned aerial systems, bridge engineering, and aerospace operations. Prof. Moreu’s projects are funded by the DOE, NSF, ONR, NAS, US DOT, TRB, and the commercial sector. He is a registered Professional Engineer since 2010. Professor Moreu is a Fellow of the American Society of Civil Engineers (ASCE) and the vice chair of the ASCE Engineering Mechanics Institute (EMI) Technical Committee in Structural Health Monitoring and Control (SHMC) (2026-2029); the chair of ASCE EMI Education Committee (2024-2027); and the vice chair of the Society for Experimental Mechanics Technical Division of Dynamics of Civil Structures (2025-2027).
The Hidden Cost of Convenience: Privacy Risks in Everyday Online Systems
Afsah Anwar, University of New Mexico
Wednesday, January 21, 2026, 2:00 PM
Location: Larrañaga Engineering Auditorium (Centennial 1041)
Abstract:
Digital service providers often prioritize frictionless user experiences by adopting technologies that simplify access to their services. One widely used mechanism is the Short Message Service (SMS), which delivers links (URLs) that enable single-click access to online services with minimal user effort. However, SMS is inherently insecure, and numerous studies and incident reports have documented message interception and data leakage. As a result, placing excessive trust in this insecure channel creates opportunities for unintended access and adversarial exploitation. This talk will present investigation techniques for identifying and analyzing the privacy implications of SMS-delivered URLs. Particularly, we will focus on Personally Identifiable Information (PII) exposures and examine their root causes, including weak authentication models and insufficient authorization controls employed by digital services. Additionally, we will explore how such PII leaks can affect users beyond those directly exposed, as well as how service-side over-fetching of sensitive data amplifies privacy risks.
Bio:
Afsah Anwar is an Assistant Professor in the Department of Computer Science at the University of New Mexico (UNM). He leads the Beyond Defense Lab, which investigates the characterization of malicious activities on the Internet and their broader societal impacts, with research interests that encompass both terrestrial and space-based systems. His research efforts are supported by the National Science Foundation (NSF) and UNM and have resulted in several recognitions, including a Best Paper Award and multiple vulnerability disclosures. He received his Ph.D. from the University of Central Florida.