137091 research outputs found
Sort by
Co-designing Robot Dogs with and for Neurodivergent Individuals: Opportunities and Challenges
No full textAutism ranks as among the most common neurodevelopmental disorders and impairs the socioemotional behaviors of those affected. Existing research has demonstrated that animals hold significant potential in impacting the socioemotional well-being of autistic individuals and their non-judgmental nature can offer comfort and support for ND individuals to be in control of their environment when animals obey their issued commands. However, much of this research has been focused in therapeutic and educational settings, leaving much room for exploration of how robot dogs can be utilized in the context of everyday interactions. With prior research demonstrating the great impacts animals can have on the lives of neurodivergent individuals and the lack of research of how these robot dogs can be integrated into their daily lives, we aim to investigate how robot dogs can be used in such a way. In addition, we also employ the use of co-design studies to integrate the feedback provided by neurodivergent individuals to ensure that the design model we envision is representative of the needs voiced by the target population.
Work in understanding these design preferences is limited. To explore more of what makes an effective design in addition to behaviors, we conduct several workshops over the course of several weeks and analyze the data. These workshops will expose neurodiverse students in Georgia Tech’s EXCEL program to two different types of robot dogs (see Figure 1), each capable of different actions, behaviors, and having different looks. Afterwards, students will participate in a participatory design process that will last over the course of 2-3 weeks. Students will split off into groups and based on their experience with the robot dogs, formulate use cases and brainstorm ideas for features and use cases they would personally like to see (see Table 1).UndergraduateComputer Scienc
Dynamic Pricing System for Physical Internet Enabled Hyperconnected Less-than-Truckload Freight Logistics Networks
No full textLess-than-truckload (LTL) shipping plays a critical role in modern supply chains by consolidating freight from multiple shippers into shared vehicles. Despite its operational flexibility and potential sustainability benefits, the LTL sector faces persistent challenges, including high per-unit costs and financial instability, as evidenced by recent industry bankruptcies. This paper investigates two structural issues limiting LTL performance: the constrained consolidation potential imposed by proprietary logistics networks, and the inefficiency of fixed pricing models that fail to reflect real-time network conditions. To address these, we explore a Physical Internet (PI)-enabled, hyperconnected LTL logistics system based on open asset sharing and dynamic flow consolidation. We then propose a dynamic pricing framework tailored for this network. Through a simulation-based study grounded in Freight Analysis Framework data and cost estimates from industry sources, we evaluate system performance across three demand and cost uncertainty scenarios in the Southeastern U.S. The results validate our system’s effectiveness and suggest a promising path forward for building more efficient LTL logistics operations
Companion Code and Dataset for Article: Bead Geometry Prediction in Wire Arc Directed Energy Deposition Using Physics-informed Machine Learning and Low-fidelity Data
No full textDataset and Code for:
Rashid, A., Vatandoust, F., Kota, A., & Melkote, S. N. (2025). Bead geometry prediction in wire arc directed energy deposition using physics-informed machine learning and low-fidelity data. Additive Manufacturing, 109, 104881.
Overview
This repository contains the full dataset and source code used in the above publication. The structure is organized into self-contained subfolders, each of which includes:
• Relevant code (model setup, model architecture, training pipeline, etc.)
• Corresponding datasets
• Trained model files
• Model outputs and generated figures (as used in the paper)
Folder Descriptions
• geometry_physics/: Contains code and data related to geometry and physics-informed modeling.
• geometry_physics_and_data/: Combines data-driven and physics-informed models
• temperature_high_fidelity/: High-fidelity thermal simulation model setup, training, visualization.
• temperature_low_fidelity/: Low-fidelity thermal simulation model setup, training, visualization.
Instructions
• Each subfolder can be run independently without modification to paths or directory structures.
• Ensure that the full dataset is downloaded and extracted.
• The Python scripts (e.g., Jupyter notebooks) are ready to run as-is.
• Required Python packages must be installed in the user's environment.
Citation
If you use this dataset or code in your own research, please cite the paper and this dataset appropriately.This repository contains the dataset and code scripts for the following journal article: Rashid, A., Vatandoust, F., Kota, A., & Melkote, S. N. (2025). Bead geometry prediction in wire arc directed energy deposition using physics-informed machine learning and low-fidelity data. Additive Manufacturing, 109, 104881
Automated extraction and synthesis of biomedical data for AI-driven systematic review and meta-analysis
No full textBiomedical literature is not simply a record of scientific discovery; it also provides a platform for research exploration and optimized clinical practice. The purpose of this thesis is to utilize and develop natural language processing methods to enhance and automate biomedical literature-based research inquiry. Specifically, we develop datasets, methods, and systems to enable AI-assisted systematic review and meta-analysis of clinical literature. We further validate its efficacy via several clinical case studies that demonstrate its value in identifying potential treatments for emerging diseases and elucidating the mechanisms by which diseases affect patients.
Qualitative systematic reviews perform a thorough survey of a particular medical topic to highlight relevant relationships and highlight promising directions for future research. To enable faster systematic review of biomedical relationships, we build a knowledge graph of relationships between biomedical entities extracted from 33+ million research articles on PubMed. We pair this with an unsupervised graph ranking algorithm that identifies related concepts and their relationships from literature. This graph and accompanying software package form a Literature Based Discovery (LBD) system that can comprehensively identify and rank disease risks, mechanisms, and repurposed drugs for future clinical or experimental research prioritization.
Similarly, quantitative meta-analysis of clinical studies forms the gold standard for establishing clinical guidelines and best practice by calculating an aggregate effect size from a collection of smaller cohorts. Meta-analysis begins with a specific research question and then extracts study-specific data elements to form a large, synthetic statistical cohort. Currently, the process of selecting research articles and extracting relevant data is done manually, taking a year on average for each clinical meta-analysis. This thesis presents data and methodological resources that dramatically accelerates the process of qualitatively and quantitatively aggregating evidence from biomedical research. In doing so, we provided the following contributions:
• We developed SemNet 2.0, a literature-based discovery software that integrates 33+ million PubMed articles into a comprehensive knowledge graph using named entity recognition, entity linking, and relationship extraction. We performed real-world case studies to illustrate the efficacy of SemNet 2.0 for summarizing relationships and prioritizing future experimental and clinical research.
• We meticulously annotated data resources -- BioSift and TrialSieve -- that enable efficient filtering of clinical studies and detailed extraction of study design and outcome information. Specifically, TrialSieve is the first dataset to our knowledge that enables the automated quantification of clinical outcomes for each group represented in a clinical study.
• We developed an interface to enable real-time, human-in-the-loop identification, filtering, and information extraction from clinical trials using large language models.
• We demonstrated the translation potential of our developed platform by creating a large database of clinical evidence for over 100 commonly used drugs with high potential to improve therapeutic outcomes for numerous types of cancer.
The deliverables of this thesis comprised seven published journal articles or conference proceedings and one under-review conference proceeding authored by David Kartchner. Specifically, this thesis included four high-quality biomedical or information science journal articles and four top-tier conference papers.Ph.D.Computational Science and Engineerin
Computational models for bacterial dynamics in community and treatment contexts
No full textMicrobes are key players in human health and disease; however, there is much debate over the nature, consequences, and importance of interactions between bacteria and their environments on the population scale. Interactions in bacterial communities and infection environments are complex and present challenges for modeling, measurement, and inference. However, rising interest in microbiomes (multi-species microbial communities), increasing antimicrobial resistance, and the quest for novel therapeutic strategies to combat human bacterial infection, all center around being able to answer common questions: how do bacteria grow and interact with each other and their environments on the population level? How do they respond to external perturbation from antibiotic exposure or bacteriophage? Using a range of mathematical approaches, we address these questions by integrating forward models and data-driven methods to assess the impacts of underlying mechanisms, abiotic and biotic perturbations, and spatio-temporal heterogeneity as they relate to microbial dynamics in human infections.
Throughout this dissertation, we employ mathematical modeling as a tool to bridge gaps between theoretical and empirical microbiology, highlighting that many standard models and inference methods fail to capture qualitative and quantitative features of microbial dynamics. First, we challenge the received wisdom that antibiotic resistance genes always worsen treatment outcomes and should be strictly minimized. We mathematically explore the effects of ecological interactions on antibiotic treatment in a two lineage system of a pathogen and commensal, proposing an optimization approach to antibiotic resistance management. We define conditions for competitive release and “beneficial” commensal resistance—namely, when commensals inhibit pathogens—and demonstrate generality to resource explicit and spatially extended models. These results are conserved in a four-species experimental community with phage, showing that the addition of phage, targeting the dominant competitor in the community, leads to extinction of the dominant species, competitive release of the next strongest competitor, and maintenance of community diversity. Next, we present an iterative approach for understanding antibiotic and inoculum effects on bacterial growth and yield. Using fine-scale experimental data and a menu of standard population models, we conclude that both growth rate and yield are modified by antibiotic exposure and that populations exhibit distinct regimes of dynamical behavior given distinct exposure conditions. Finally, we expand our modeling into two-dimensional space, building an agent-based simulation of bacterial cells and aggregates to explore physical and socio-microbiology mechanisms underlying relationships between bacterial growth rate and aggregate size.
This work has important implications for both theoretical and empirical studies of microbial systems—evaluating and informing methods for sampling, inference, and modeling to efficiently capture underlying complexities of interactions between bacteria and their environments. In the study of human infection, we provide a baseline toolkit to develop improved treatment strategies for acute and chronic infections and to increase predictability of treatment outcomes.Ph.D.Quantitative Bioscience
Multi-Sensor Tasking for Ground-Based Space Situational Awareness via Job-Shop Scheduling Problem
No full textSensor-tasking to collect observation measurements of resident space objects constitutes a central problem for Space Situational Awareness (SSA). One specific type of sensor-tasking problem aims at rapidly producing an optimal or near-optimal schedule spanning over an observation horizon of a few hours up to a few nights. We derive tractable binary linear program formulations for solving the scheduling problem for both single and multiple telescopes. Observations are modeled by assuming the telescopes must track orbit passes for a user-defined exposure time. The proposed programs are efficiently solvable by branch-and-bound algorithms for instance sizes of practical applicability
Reference-frame dependent differences in spatial navigation network functional connectivity
No full textSpatial navigation deficits are an early indicator of aging-related cognitive decline. Successful navigation depends on integrating allocentric and egocentric reference frames (RFs), processed by the hippocampus and posterior parietal cortex (PPC). The retrosplenial cortex (RSC) shares projections with both regions, facilitating RF integration through multimodal sensory processing. However, the impact of aging and RF preference on network connectivity remains unclear. We measured PPC-RSC functional connectivity using functional magnetic resonance imaging (fMRI) in younger adults during a Y-Maze navigation task. Participants were classified as having an allocentric or egocentric RF preference based on their performance during the Y-Maze probe trials. Voxelwise and region-based functional connectivity analyses were conducted to assess measures of RSC-PPC connectivity during resting-state and task-based conditions. Results revealed distinct RSC-PPC connectivity patterns when comparing navigation preferences across scanning conditions. Participants with an egocentric RF preference demonstrated greater resting-state RSC-PPC connectivity, whereas participants with an allocentric RF preference demonstrated greater task-based RSC-PPC connectivity. These findings indicate context-dependent activation within the spatial navigation network and demonstrate the impact of individual RF preferences and biases on network connectivity patterns.UndergraduateNeuroscienc
President's Newsletter
No full textA few weeks ago, I traveled to South Korea with some colleagues to visit our partners at Hyundai Motor Group. The trip left me extremely excited for the future of our state’s economy in the global marketplace and proud of Georgia Tech’s role in it