AIM Seed Grants Support 22 AI Research Projects

From developing futuristic packaging that guards food freshness and safety, to improving speech therapy for neurodiverse people, to developing novel sea ice monitoring techniques, 22 new artificial intelligence (AI) research projects were awarded through a new University of Maryland seed grant program, totaling about $1.5 million.

The grants are administered by the Artificial Intelligence Interdisciplinary Institute at Maryland (AIM), a collaborative hub that the university launched last spring to conduct research, offer innovative and experiential learning opportunities for students, and focus on responsible and ethical AI technology to advance the public good. UMD and its philanthropic and industry partners plan to invest more than $100 million in the institute over the next 10 years.

"This year’s AIM research seed awards fund bold, interdisciplinary projects that embody the institute’s mission: to advance innovative artificial intelligence research through our four pillars of accessibility, social justice, sustainability and learning,” said Sheena Erete, AIM associate director of research and College of Information associate professor. “We are proud to support scholars who are expanding what’s possible through thoughtful, inclusive and impactful research.”

The grants were awarded in several categories, with funding from AIM and matching funds from campus units.

Cross-College Collaborative Awards ($100,000-$300,000)

AI and Robotics Accelerated Discovery of Biobased, Sustainable, Antimicrobial Food Packaging for Enhanced Postharvest Safety and Security
Chemical and Biomolecular Engineering Assistant Professor Po-Yen Chen, Chemistry and Biochemistry Assistant Professor Mercedes Taylor and Nutrition and Food Science Professor Qin Wang plan to integrate AI-driven predictive modeling with robot-automated, high-throughput experimentation to accelerate the development of next-generation bio-based food packaging.

CommunityTwin: A Digital Twin for Enhanced Decision-Making in Public Health
College of Information Associate Professor Vanessa Frías Martínez; Dean’s Professor of Information Systems and UMIACS Professor Louiqa Raschid; Linguistics and UMIACS Professor Philip Resnik; Decision, Operations and Information Technologies Assistant Professor Eaman Jahani; and University of Washington Associate Professor of Health Systems and Population Health Neil Sehgal plan to develop a highly detailed yet adaptable set of models of a community known as a “digital twin” to assist policymakers in designing personalized, localized interventions for pressing community challenges. The system will integrate deep learning, human mobility data and fine-tuned large language models to simulate real-world behaviors and community mental models.

AI-Driven Sensor Fusion for Arctic Sea Ice Mapping
Art Associate Professor Cy Keener, Geographical Sciences and Atmospheric and Oceanic Science Associate Professor Sinead Farrell, and Computer Science and UMIACS Assistant Professor Christopher Metzler plan to develop advanced AI-driven photogrammetry and sensor fusion techniques to enhance long-term coastal monitoring efforts. They will identify scientifically significant Arctic sea ice regions and then develop low-cost and accessible sensing platforms to survey the areas and form high-resolution 3D reconstructions of the ice above and below the sea surface.

Human-Centered AI for Translating Classical Chinese
Computer Science and Institute for Advanced Computer Studies (UMIACS) Associate Professor Marine Carpuat and School of Languages, Literatures, and Cultures Associate Professor Andrew Schonebaum plan to design human-centered AI technology to reliably translate classical Chinese and integrate AI literacy into the humanities curriculum.

AI for Motor Learning in Instrumental Education
UMIACS Research Scientist Cornelia Fermüller and Music Professor Irina Muresanu plan to develop AI-driven tools to help students master bow techniques for string instruments. They will develop algorithms that analyze video and audio inputs to provide precise, actionable feedback to students.

Theory and Metrics for Law-Informed Attribution of Human-Generative AI Liability
College of Information Assistant Professor Zubin Jelveh, Computer Science and UMIACS Associate Professor Furong Huang and Computer Science graduate student Tin Nguyen plan to develop a framework to determine how much liability companies that deploy AI systems should bear compared with users who may be harmed by the technology. The framework will be informed by the “joint and several liability” principle from tort law and the “path model of blame” from psychology.

Blind Motion Prediction for Robotic Planning
College of Information Associate Professor Hernisa Kacorri and Computer Science and UMIACS Associate Professor Abhinav Shrivastava plan to collect and share a novel multimodal motion benchmark dataset tailored to the movements of individuals who are blind. Using this dataset, they will evaluate the ability of generative AI models to predict how blind individuals interact with objects and assess their potential in developing robots that can interact more safely and smoothly with visually impaired people.

AI-Driven Development of Neurodiversity Affirming Practices for Teachers and Clinicians (ADAPT)
For this community-based grant, Counseling, Higher Education and Special Education Assistant Professor Veronica Kang; Hearing and Speech Sciences Assistant Professor José Ortiz; and Psychology Assistant Research Professor Heather Yarger will collaborate with Kate Lu, clinical director and parent representative at the Chinese Culture and Community Service Center Health Clinic, and Ariel La, a speech language pathologist at Speech Gardens, to train UMD students to use the TerpAI chatbot in conjunction with ADAPT. This checklist for neurodiversity-affirming and culturally sustaining instruction supports emotion regulation, executive functioning, communication, and social and play skills for culturally diverse autistic youths in Maryland and Virginia.

Intergenerational DetAIctives: Connecting Older Adults and Undergraduates for Community-Driven Generative AI Exploration
College of Information Associate Professor Amanda Lazar, Assistant Professor Caro Williams-Pierce and Associate Professor Tamara Clegg plan to generate insights in generative AI design and use through intergenerational, equitable learning in community-driven informal learning systems. They will partner with two community-based learning organizations in Prince George’s County that serve minoritized groups to teach participants to critically consider a variety of uses for AI, learn tips for identifying AI products, and use various forms of generative AI as powerful but imperfect tools that can support them in accomplishing life goals.

Participatory Design of AI Tools for Neurodiversity-Affirming Speech Therapy
Maryland Language Science Center Assistant Research Professor Shevaun Lewis and Hearing and Speech Sciences Clinical Professor Kathy Dow-Burger plan to develop a new framework for collaborative analysis and goal setting between speech-language therapists and their clients that is centered around objective conversation data. They plan to engage autistic adults and speech-language therapists in the participatory design of AI tools for analyzing recordings of everyday conversations.

Co-Learning Code and Mind: Integrating AI With Transformative Social and Emotional Learning for Diverse Youth Through Tangible Robotics
For this community-based grant, Counseling, Higher Education, and Special Education Associate Professor Chunyan Yang and Computer Science and UMIACS Assistant Professor Huaishu Peng plan to engage middle school students alongside their neurodiverse peers in the hands-on co-designing, co-building and co-programming of robots. They will partner with two nonprofit organizations in Montgomery County for this project: the AOE Robotics Club and the Special Education Equal Development Society.

Individual Faculty Awards ($5,000-$20,000)

Effects of Human-Robot Collaboration on Learning of Complex Action Sequences
Kinesiology Associate Professor Rodolphe Gentili plans to investigate whether an expert-level humanoid robot with adaptive planning abilities that jointly performs a complex cognitive-motor task with a novice human teammate improves human learning. This work will inform the development of smart assistants that help people with cognitive planning disabilities.

A Data-Driven Deep Learning Strategy for High-Performance Atomic Catalyst Design
Mechanical Engineering Professor Teng Li plans to develop a data-driven, deep learning-accelerated design methodology to speed up the discovery of high-performance atomic catalysts by efficiently screening compositions and structures. The framework could significantly reduce the time and cost associated with materials discovery in clean energy, sustainable manufacturing and national security applications.

Project MELD: AI-Driven Scaffolding for Scientific and Civic Reasoning
Human Development and Quantitative Methodology Professor Doug Lombardi plans to develop and test an AI-driven instructional tool that deepens scientific and civic reasoning in K-12 education. The tool will help students critically evaluate complex scientific issues and support teachers in fostering deeper learning, critical thinking and analytical reasoning.

Leveraging AI to Analyze Public Framing and Perceptions of AI
College of Information Assistant Professor Julia Mendelsohn plans to develop computational methods to analyze the framing of AI on social media and conduct survey experiments to measure the impact of framing on attitudes toward AI. This research aims to understand public hopes and concerns about AI, explain why some people embrace AI while others strongly oppose it and assess how online portrayals of AI shape public attitudes.

How Far Is Fair? Measuring Legal Boundaries in AI-Generated Content
Decision, Operations and Information Technologies Assistant Professor Lauren Rhue plans to examine an overlooked question: How different does AI-generated content need to be from original copyrighted sources to qualify as fair use rather than derivative work? Rhue plans to systematically generate AI text in the style of specific writers and their original works, acquire expert opinions on the similarity between the AI-generated and original texts, develop a comprehensive distance measure using various similarity metrics, validate this measure through experiments with IP experts and law students, and fine-tune large-language models to identify copyright violations. This work may benefit less-established writers, who can struggle to enforce their copyright due to resource constraints.

Individual Postdoc and Student Awards (up to $5,000)

Introducing Spatial Speech Context in Wearable Large Language Models
Computer Science graduate student Ayushi Mishra, who is advised by Assistant Professor and Computer Science and UMIACS Assistant Professor Nirupam Roy, plans to develop a novel system architecture that incorporates spatial speech understanding into large-language models, enabling contextually aware and adaptive applications for wearable technologies. Spatial speech integration elevates wearables into intelligent companions—enhancing productivity, safety and convenience in many scenarios.

Evaluation of the Use of AI Voices in the TOEFL Junior Standard Test Listening Section
School of Languages, Literatures, and Cultures graduate student Sanshiroh Ogawa, who is advised by Assistant Professor Bronson Hui, plans to scrutinize the use of synthetic speech in language tests by using listening comprehension items from the TOEFL Junior Standard test recorded by human voice actors and AI-based and non-AI-based text-to-speech systems. The project will analyze the test performance of learners of English in secondary school in Japan to determine whether AI-based and human-based listening assessment items are comparable.

Generative AI and Online Information-Seeking Practices
College of Information graduate student Maithreyi Pejathaya, who is advised by Assistant Professor Caro Williams-Pierce, will explore online information-seeking processes among AI novices to better understand how individuals decide when and how to use and trust generative AI information-seeking tools. Additionally, they will investigate how novices evaluate the accuracy of generative AI-produced text-based content compared with experts.

Multimodal Volumetric Imaging and Material Identification of Multilayered Objects
Computer Science graduate student Harshvardhan Takawale, who is advised by Computer Science and UMIACS Assistant Professor Nirupam Roy, plans to develop a multimodal volumetric imaging framework that integrates millimeter-wave radar and ultrasonic transducers for high-fidelity 3D reconstruction and material identification of multilayered heterogeneous objects. Their work will address critical focus areas in AI-driven sensing, non-destructive testing and medical imaging, where accurate subsurface inspection or complex material characterization is vital.

Artificial Intelligence-Aided Causal Discovery of Failures in Hydrogen Fueling Stations
Mechanical Engineering Postdoctoral Associate Ruochen Yang, who is advised by Associate Professor Katrina Groth, plans to leverage operational data from hydrogen fueling stations to help uncover dependencies unknown to human operators and designers, optimize system performance and detect failure patterns that may not be immediately apparent. They expect to gain deeper insights into the causal relationship between key operational variables at hydrogen fueling stations, ultimately enhancing their safety, efficiency and reliability.

AIM Fellow Program

Establishing an AI Technology Policy Hub of International Standing at UMD
College of Information and UMIACS Professor Katie Shilton and School of Public Policy Associate Research Professor Charles Harry plan to bring Lee Tiedrich, an internationally renowned scholar of AI, to UMD for the one-year AIM Fellow residency. She will engage with the College of Information’s Tech Policy Hub (led by Ido Sivan-Sevilla), which encourages policy practitioners, industry leaders and AI governance scholars from the D.C. metro area to connect and share knowledge on the design, implementation and enforcement of AI policies at all levels.