Electrical and Computer Engineering Commons^™

One Size Does Not Fit All: Revisitingworld Models And Neurosymbolic Ai, Amit P. Sheth, Madhur Thareja, Anushka Pawar, Niyati Rawal

Publications

World models are being built twice, from opposite ends, without a shared theory of how the two halves should meet. One lineage grounds the world model in perception: a self-supervised, latent-predictive encoder – exemplified by Joint Embedding Predictive Architectures (JEPA) – that learns the structure of sensory experi-ence. A second, older lineage grounds the world model in cognition: an explicit, inspectable structure of entities, rules, and constraints, ranging from knowledge graphs to formal logic to physical law. Neither lineage alone has produced a world model that is simultane-ously adaptive and auditable. We argue this is not solved by picking a …

AiXGa1−XN (0.7 < X < 1) Schottky Diodes Using Distributed Polarization Doped Layer With A Current Density Of 14 Ka/Cm2, Tariq Jamil, Abdullah Al Mamun Mazumder, Mafruda Rahman, Muhammad Ali, Grigory Simin, M. Asif Khan

Faculty Publications

High Al-content Al_xGa_1-xN (0.7 < x < 1) quasi-vertical Schottky barrier diodes (SBDs) with distributed polarization doping were grown on the bulk AlN substrate. They exhibit excellent rectification behavior with a large forward current density (~14 kA/cm² ) and a high breakdown field of ~8.3 MV/cm. The SBDs also exhibited low ideality factors of (n~1.2) with a high Schottky barrier height (Φ _b~ 1.7 eV). Thus, this study demonstrates the feasibility of the distributed polarization doping approach for high current–high voltage devices.

Hybrid Cnn–Gru-Based Demand–Supply Forecasting To Enhance Sustainability In Renewable-Integrated Smart Grids, Süleyman Emre Eyimaya, Necmi Altin

Faculty Publications

The rapid integration of renewable energy sources in smart grids has introduced significant uncertainty in both power generation and consumption patterns, posing challenges to environmental, economic, and operational sustainability. Accurate short-term forecasting of energy demand and supply is essential for achieving optimal scheduling, grid stability, and resilient operation in renewable-integrated power systems. This study proposes a hybrid deep learning framework combining Convolutional Neural Networks (CNN) and Gated Recurrent Units (GRU) for intelligent joint demand–supply forecasting in smart grids. The model was developed and implemented in MATLAB using real-world datasets comprising electricity consumption, photovoltaic (PV) generation, temperature, and irradiance variables. Comparative …

Biologically-Inspired Multiscale Neuromorphic Architecture, Christian O'Reilly, Ramtin Zand

Publications

This white paper proposes a biologically-inspired multiscale neuromorphic architecture that bridges key gaps between artificial neural networks (ANNs), spiking neural networks (SNNs), and biological neural networks (BNNs). While SNNs offer promising energy efficiency, their broader adoption remains limited by suboptimal performance and the need for novel learning paradigms. To address these challenges, the proposed framework integrates structural and functional principles observed in the brain, including hierarchical organization, sparse and modular connectivity, predictive coding, and diverse neuronal dynamics.

The architecture operates across micro-, meso-, and macro-scales, incorporating neuron-level diversity (e.g., excitatory/inhibitory and principal/support cells), canonical microcircuits (CMCs), and large-scale hierarchical organization. …

In-Situ Eval: A Modular Framework For Custom And Real-Time Rag Benchmarking, Ritvik Garimella, Kaushik Roy, Chathurangi Shyalika, Amit Sheth

Publications

Retrieval-Augmented Generation (RAG) has become the standard approach for integrating domain knowledge into Large Language Models (LLMs). However, fair comparison of RAG pipelines remains difficult: data preparation is often ad hoc, subsampling methods are opaque, parameters vary across implementations, and evaluation is fragmented. We present In-Situ Eval, a unified and reproducible framework that operationalizes the full RAG pipeline with configurable subsampling strategies and both RAG-specific and generic evaluation metrics. The platform supports two execution modes: an offline Dataset mode for evaluating precomputed outputs, and a live Retrieval mode for benchmarking RAG variants with state-of-the-art LLMs. Users can flexibly select datasets, …

Temperature Dependent High Frequency Performance Of A 62% Algan Channel Hemt, Jiahao Chen, Abdullah Al Mamun Mazumder, Parthasarathy Seshadri, Dheekshinn Nandakumar, Ruixin Bai, Rafael Andrew Choudhury, M. Asif Khan, Chirag Gupta

Faculty Publications

This article reports on the temperature dependent performance of a HEMT with an Al_0.62Ga_0.38N channel layer and an Al_0.84Ga_0.16N barrier layer grown by metal–organic chemical vapor deposition. The device in this report was measured at room temperature and elevated temperatures of 100–150 °C. The sheet resistance increased from 3.5 kΩ/sq (25 °C) to 5.4 kΩ/sq (150 °C), while the contact resistance remained nominally similar. For a device with 160 nm gate length and 2 µm source-to-drain length, excellent electrical characteristics have been achieved when the device was operated at 150 °C with …

Extreme Bandgap Al0.63Ga0.37N Quasivertical Schottky Barrier Diodes With A High Baliga Figure Of Merit Of 630 Mw Cm−2, Abdullah Al Mamun Mazumder, Shahab Mollah, Kamal Hussain, Abdullah Mamun, Seongmo Hwang, Tariq Jamil, Grigory Simin, Asif Khan

Faculty Publications

"Extreme bandgap n-Al_0.63Ga_0.37N quasi-vertical Schottky barrier diodes (SBDs) with doping densities of ≈8 × 10¹⁷cm⁻³ (Sample A) and ≈2 × 10¹⁷cm⁻³ (Sample B) are grown on an AlN/sapphire substrate using metalorganic-chemical vapor deposition (MOCVD). Sample A achieves a high forward current density of ≈59.5 kA cm⁻² at 10 V with an I_ON/I_OFF ratio of ≈10⁸ (calculated from the forward current at + 3.8 V and the reverse current at −1 V) and an ideality factor of 2.8. Sample B has a forward current …

Extreme Bandgap Recessed-Gate Metal Oxide Semiconductor Heterostructure Field Effect Transistors With Drain Current 0.28 A Mm−1 And Threshold Voltage −1.5 V, Abdullah Al Mamun Mazumder, Abdullah Mamun, Kenneth Stephenson, Kamal Hussain, Tariq Jamil, Grigory Simin, Asif Khan

Faculty Publications

Herein, the first demonstration of hybrid high-k oxide (ZrO2-Al2O3) incorporation into extreme bandgap (EBG) Al_0.87Ga_0.13N/Al_0.64Ga_0.36N metal-oxide-semiconductor heterostructure field-effect transistors (MOSHFETs) is presented, with both planar and recessed-gate designs on the same AlN/sapphire template with a state-of-the-art low contact resistance of 1.4 Ω mm (contact resistivity, ρc ≈ 5.7 × 10−6 Ω cm2). The recessed-gate MOSHFETs achieve a threshold voltage shift of ΔVTH = 5.8 V, highlighting improved channel control. Static output measurements reveal a peak drain current (IDS) of 340 mA mm⁻¹ for the planar …

Microgrid Black Start Challenges: The Role Of Grid-Forming Inverters, Gab-Su Seo, Wenzong Wang, Behrooz Mirafzal

Faculty Publications

Grid-forming (GFM) inverters are becoming increasingly important for future power systems, particularly in establishing and restarting microgrids after blackouts. The use of GFM inverters enables microgrids to operate independently of utility power and provide key advantages over synchronous generators in black start scenarios, including rapid startup and stable voltage and frequency support for critical loads. However, inverter-driven black start introduces unique challenges and operational considerations. This article examines key challenges and solutions, emphasizing inverter design, control strategies, and microgrid system requirements. Drawing on analysis, simulation, and experimental results, this article highlights the central role of GFM inverters in ensuring reliable …

Crisis Observatory: Extracting Credible Signals During A Crisis In The Age Of Llms, Kuan-Chieh Lo, Pranav Maneriker, Sriram Sai Ganesh, Dominik Winecki, Kelly Garrett, Ayaz Hyder, Arnab Nandi, Valerie Shalin, Shannon A. Bowen Ph.D., Amit Sheth, Srinivasan Parthasarathy

Publications

Systems for crisis response have required several different models for the analysis of unstructured text, such as identifying needs, locations, topics, routing, and matching of needs with available responders. Large Language Models (LLMs) have replaced task-specific models across various language processing tasks. However, LLMs are known to be limited by their training data, collected before the crisis. In this demo, we explore the use of LLMs for crisis response scenarios with rapidly evolving information environments. We show how the augmentation of these models with external reliable sources of crisis-specific information can help build adaptive systems for response. The demonstration video …

Spectral Modeling Of Electromagnetic Radiation From Power Converters, Regan A. Varner

Theses and Dissertations

Switching power converters are increasingly utilized in contemporary electronics. They are becoming more compact and operate at higher frequencies, which contributes to radiated electromagnetic interference (EMI). This EMI is important to quantify for electromagnetic compatibility of the converters with other nearby systems. In this thesis, analytical models have been developed for both ideal and non-ideal switching waveforms, employing piecewise functions to describe time-domain waveforms, with their corresponding frequency spectra derived and compared to empirical spectra. The results demonstrate that the comparisons with measured frequency data exhibit agreement with the analytical models up to approximately 25 MHz. Additionally, our analysis indicates …

Towards Robust Bmocz Schemes For Non-Coherent Wireless Communication, Anthony Joseph Perre

Theses and Dissertations

Non-coherent communication has emerged as a promising approach to address several challenges at the physical layer. A particular scheme called binary modulation on conjugate-reciprocal zeros (BMOCZ) uses the zeros (i.e, roots) of a polynomial to convey information bits. In this thesis, we strive to improve the reliability of BMOCZ in various scenarios. In particular, we utilize machine learning (ML) methods to determine the parameters for BMOCZ and improve upon the decoding performance. Moreover, we introduce a smooshed BMOCZ variant to combat typical impairments encountered within wireless communications, including timing offsets (TOs) for noncoherent orthogonal frequency division multiplexing (OFDM). Through numerical …

Addressing Key Challenges In Mhz Power Converters, Aqarib Hussain

Theses and Dissertations

Power electronics are essential in applications such as aerospace, electric vehicles, and electric ships where compact, efficient, and high power-density converters are increasingly demanded. This push toward miniaturization has led to converter designs operating at switching frequencies in the megahertz (MHz) range. Higher-frequency operation enables significant reductions in passive component sizes—particularly transformers and inductors—thereby facilitating dense packaging. However, these benefits come with new challenges, including increased thermal stress, complex semiconductor behavior, higher-frequency magnetic design constraints, and elevated electromagnetic interference (EMI).

This research addresses several key challenges in MHz converter design. The first part involves the development of a 1-MHz, 1-kW …

Generalized Dynamic Pole Placement Control For Voltage Source Converters With Inductive-Capacitive Output Filters, Byunghee Moon

Theses and Dissertations

Voltage Source Converters (VSCs) with inductive-capacitive (LC) filters are prevalent in power electronics applications. There are many control techniques utilized for that category of power converter topology. However, as converter applications continue to increase their switching frequency, faster control methods are desirable to take advantage of the higher bandwidth that increased switching frequency allows for

Dynamic pole placement (DPP) controls are very promising methods for achieving high-speed control of VSCs with LC filters. However, the conventional DPP controls have very high sensitivity to disturbances and noise. In order to address this, two novel DPP current control methods are proposed and …

Optical And Thermal Characterization Of Vertically Conducting Β-Ga₂O₃ Diodes Grown On 4h-Sic Substrates For Short-Wavelength (<254 Nm) Detection, Muhammad Hassan Tahir

Theses and Dissertations

The advancement of compact, thermally stable, and efficient devices for deepultraviolet (DUV) detection below 254 nm is critical for solar-blind sensing, radiation monitoring, and high-power electronics. This thesis presents the growth, fabrication, and detailed optical and thermal characterization of vertically conducting β-gallium oxide (βGa₂O₃) Schottky barrier diodes (SBDs) grown on n-type 4H-Silicon Carbide (4H-SiC) substrates using metal-organic chemical vapor deposition (MOCVD).

β-Ga₂O₃ is an ultra-wide bandgap (UWBG) semiconductor (~4.8 eV) with intrinsic DUV transparency and high breakdown electric field, making it an ideal candidate for short-wavelength photodetection. However, its low thermal conductivity limits its vertical device performance. To overcome this, …

Effect Of Growth Conditions On The Performance Of Vertically Conducting Beta-Ga2o3 Diodes On 4h-Sic Substrate By Mocvd, Nifat Jahan Nipa

Theses and Dissertations

This thesis highlights the critical influence of doping strategies, specifically co-delta doping with silicon (Si) and indium (In), on improving the structural quality and device performance of β-Ga₂O₃-based Schottky barrier diodes (SBDs). To achieve better thermal management, β-Ga₂O₃ thin film was grown on bulk 4H-SiC substrates using the Metal-Organic Chemical Vapor Deposition (MOCVD) technique, renowned for its precise control over composition, thickness and doping profile. The ultra-wide bandgap (UWBG), high critical breakdown field and low turn-on resistance of β-Ga₂O₃ make it a strong candidate for next-generation power electronics and short-wavelength detection in extreme environments. However, the presence of structural defects …

From Assembly Lines To The Open Road: Predicting Rare Events In Autonomous Systems, Ruwan Wickramarachchi

Publications

In the age of embodied AI and smart automation, autonomous agents are increasingly deployed in high-stakes, real-world environments. Ensuring the robustness and resilience of these systems in the face of rare but critical failure events is essential for their safe and reliable operation. Accurate forecasting of such rare events is particularly crucial, as a single overlooked anomaly can lead to catastrophic consequences. In manufacturing, for instance, unplanned downtime due to rare failures costs industries over \$50 billion annually, with sectors like automotive losing more than \$2 million per hour—even with preventive maintenance systems in place.

However, the extreme rarity and …

High Field Performance Of Si-Doped N-Aln Layers Grown Using Pulsed Mocvd, Abdullah Al Mamun Mazumder, Tariq Jamil, Mafruda Rahman, Muhammad Ali, Grigory Simin, Asif Khan

Faculty Publications

We report on the study of high-field performance of Si-doped n-AlN layers that were grown using a pulsed metalorganic chemical vapor deposition (PMOCVD) process. In the past we showed this pulsed doping approach to lead to doping efficiency superior to that in the conventional MOCVD process. Here using them as the drift layer for a quasi-vertical conduction Schottky barrier, we show their ability to withstand high reverse bias voltages and sustain an electrical field as high as 9.9 MV cm−1. Our study thus demonstrates the viability of the PMOCVD growth and doping approach to yield n-AlN layers suitable …

Extreme Bandgap Polarization Doped Algan Layers On Bulk Aln For Pn-Diodes With An 8.5 Mv Cm−1 Breakdown Field And Forward Current Density Exceeding 20 Ka Cm−2, Tariq Jamil, Abdullah Al Mamun Mazumder, Muhammad Ali, Mafruda Rahman, Kenneth Stephenson, Grigory Simin, Asif Khan

Faculty Publications

In this paper we present a study of distribution polarization doped Al_xGa_1−xN layers and their use in quasi-vertical configuration pn-diodes which exhibited a high breakdown field of ∼8.5 MV cm⁻¹ and a large forward current density (∼23 kA cm⁻²). We also establish their potential use in UVC light emitters by studying the optical emission from a quantum well inserted at the distribution polarization doped pn-junction interface.

Multiagent Copilot In Industrial Ai Applications, Chathurangi Shyalika, Renjith Prasad, Utkarshani Jaimini, Cory Henson, Fadi El Kalach, Amit Sheth

Publications

In the era of smart automation and digital transformation, achieving efficiency, precision, and adaptability is essential for industries to remain competitive. Sectors, including manufacturing, supply chain and logistics, healthcare, finance, and retail, face significant challenges in deploying Artificial Intelligence (AI) solutions tailored to their unique needs, particularly in critical, resource-constrained applications. According to Gartner’s 2024 Hype Cycle for Artificial Intelligence, composite AI, which integrates techniques like machine learning, knowledge graphs, and rule-based systems, is becoming foundational for industries, enhancing predictions, decisions, and scalability across complex environments.

The complexity of real-world systems requires Industrial AI solutions to be customizable to business …

Scalable Cyber-Physical Testbed For Cybersecurity Evaluation Of Synchrophasors In Power Systems, Shuvangkar Chandra Das, Tuyen Vu, Hebert L. Ginn Iii

Faculty Publications

This paper presents a synchrophasor-based real-time cyber-physical power system testbed with a novel security evaluation tool, pySynphasor, that can emulate different real attack scenarios on the phasor measurement unit (PMU). The testbed focuses on real-time cyber-security emulation using different components, including a real-time digital simulator, virtual machines (VM), a communication network emulator, and a packet manipulation tool. The script-based VM deployment and software-defined network emulation facilitate a highly scalable cyber-physical testbed, which enables emulations of a real power system under different attack scenarios such as address resolution protocol (ARP) poisoning attack, man-in-the-middle (MITM) attack, false data injection attack (FDIA), and …

Examining Physiological Responses To Misophonic Triggers, Christian O'Reilly, Xuan Yang, Sewon Oh, Doug Wedell, Svetlana Shinkareva

Faculty Publications

We collected and analyzed an array of biosignals (face electromyogram, skin electrodermal activity, peripheral temperature, and electrocardiogram) in 60 participants with and without misophonia, a condition characterized by decreased tolerance to innocuous sounds. Our goal was to objectively characterize the physiological response to misophonia triggering sounds. We found that misophonic responses can be objectively identified in some cases through atypical physiological reactions to triggering stimuli, though not all participants exhibited this response. Our analyses suggest a large interindividual variability in response to misophonic triggers and highlights the need for methodological adjustments in future experiments to increase the detectability of misophonic …

Removing Eog Artifacts From Eeg Recordings Using Deep Learning, Christian O'Reilly, Scott Huberty

Faculty Publications

The electroencephalogram (EEG) directly measures the electrical activity generated by the brain. Unfortunately, it is often contaminated by various artifacts, notably those caused by eye movements and blinks (EOG artifacts). Such artifacts are usually removed using an independent component analysis (ICA) or other blind source separation techniques. However, it is difficult to assess whether subtracting EOG components estimated through ICA removes some neurogenic activity. It is crucial to address this question to avoid biasing EEG analyses. Toward that objective, we developed a deep learning model for EOG artifact removal that exploits information about eye movements available through eye-tracking (ET). Using …

A Reliable And Efficient Detection Pipeline For Rodent Ultrasonic Vocalizations, Sabah Shahnoor Anis, Devin Mark Kellis, Kris Ford Kaigler, Marlene A. Wilson, Christian O'Reilly

Faculty Publications

Analyzing ultrasonic vocalizations (USVs) is crucial for understanding rodents' affective states and social behaviors, but the manual analysis is time-consuming and prone to errors. Automated USV detection systems have been developed to address these challenges. Yet, these systems often rely on machine learning and fail to generalize effectively to new datasets. To tackle these shortcomings, we introduce ContourUSV, an efficient automated system for detecting USVs from audio recordings. Our pipeline includes spectrogram generation, cleaning, pre-processing, contour detection, post-processing, and evaluation against manual annotations. To ensure robustness and reliability, we compared ContourUSV with three state-of-the-art systems using an existing open-access USV …

Swvl: A Custom Ai-Powered Face Tracking Camera Gimbal, Alexander J. Anderson-Mcleod, Jakub Jerzmanowski, Michael Laitarovsky, Trevor Allison, Jagger Tanner

Senior Theses

In response to the growing demand for smarter, more responsive face tracking cameras in the post-pandemic world, our team designed SWVL, a custom AI-powered face tracking gimbal meant to address the limitations commonly encountered by the commercial models currently on the market. These commercially available gimbals come with several issues, such as frequently losing track of the person in the frame and requiring manual resets, which we sought to fix with our implementation. We designed a system with fully custom hardware and software including a 3D printed dual-axis camera gimbal driven by stepper motors, a control PCB based around an …

A Digital Twin Based Forecasting Framework For Power Flow Management In Dc Microgrids, Kerry Sado, Jarrett Peskar, Austin Downey, Jamil A. Khan, Kristen Booth

Faculty Publications

The ability to forecast system conditions is integral to the definition and functionality of digital twins. While forecasting methods have been explored for use in digital twin systems, the integration of feedback mechanisms for real-time forecasting and in-situ decision-making in DC microgrids has not been extensively investigated. This research develops a modular forecasting framework tailored for digital twins in DC microgrids to enable real-time monitoring, online forecasting, and decision-making. DC microgrids, characterized by dynamic load variations, benefit from advanced predictive capabilities to maintain stability and operational efficiency. The proposed digital twin-based forecasting framework addresses these challenges by providing real-time predictive …

Si-Doped Ain Using Pulsed Metalorganic Chemical Vapor Deposition And Doping, Tariq Jamil, Abdullah Al Mamun Mazumder, Mohammod Ali, Jingyu Lin, Hongxing Jiang, Grigory Simin, M. Asif Khan

Faculty Publications

In this paper we describe a pulsed metalorganic chemical vapor deposition (MOCVD) Si-doping approach for AlN epilayers over bulk AlN. The Al-rich growth/doping conditions in the pulsed MOCVD process resulted in n-AlN layers with transmission line model currents that were an order higher than for structures on layers that were grown/doped at identical temperatures using the conventional MOCVD process. Our work demonstrated that like the other reported approaches such as UV exposure during growth, the pulsed MOCVD process is also very effective in reducing point defects by the defect quasi-Fermi level-chemical potential control.

2 Kv Al0.64Ga0.36N-Channel High Electron Mobility Transistors With Passivation And Field Plates, Md Tahmidul Alam, Jiahao Chen, Kenneth Stephenson, Md Abdullah-Al Mamun, Abdullah Al Mamun Mazumder, Shubhra S. Pasayat, Asif Khan, Chirag Gupta

Faculty Publications

High voltage (∼2 kV) Al_0.64Ga_0.36N-channel high electron mobility transistors were fabricated with an on-resistance of ∼75 Ω. mm (∼21 mΩ. cm²). Two field plates of variable dimensions were utilized to optimize the breakdown voltage. The breakdown voltage reached >3 kV (tool limit) before passivation however it reduced to ∼2 kV after Si₃N₄ surface passivation and field plate deposition. The breakdown voltage and on-resistance demonstrated a strong linear correlation in a scattered plot of ∼50 measured transistors. The fabricated transistors were electrically characterized and benchmarked against the state-of-the-art high-voltage (> 1 kV) …

A Survey On Food Ingredient Substitutions, Hyunwook Kim, Revathy Venkataramanan, Amit P. Sheth

Publications

Diet plays a crucial role in managing chronic conditions and overall well-being. As people become more selective about their food choices, finding recipes that meet dietary needs is important. Ingredient substitution is key to adapting recipes for dietary restrictions, allergies, and availability constraints. However, identifying suitable substitutions is challenging as it requires analyzing the flavor, functionality, and health suitability of ingredients. With the advancement of AI, researchers have explored computational approaches to address ingredient substitution. This survey paper provides a comprehensive overview of the research in this area, focusing on five key aspects: (i) datasets and data sources used to …

Exploring The Potential Of Large Language Models For Assisting With Mental Health Diagnostic Assessments: The Depression And Anxiety Case, Kaushik Roy, Harshul Surana, Darssan Eswaramoorthi, Yuxin Zi, Vedant Palit, Ritvik Garimella, Amit Sheth

Publications

Large language models (LLMs) are increasingly attracting the attention of healthcare professionals for their potential to assist in diagnostic assessments, which could alleviate the strain on the healthcare system caused by a high patient load and a shortage of providers. For LLMs to be effective in supporting diagnostic assessments, it is essential that they closely replicate the standard diagnostic procedures used by clinicians. In this paper, we specifically examine the diagnostic assessment processes described in the Patient Health Questionnaire-9 (PHQ-9) for major depressive disorder (MDD) and the Generalized Anxiety Disorder-7 (GAD-7) questionnaire for generalized anxiety disorder (GAD). We investigate various …