In today’s research landscape, solving complex challenges requires interdisciplinary collaboration. At ISI, researchers from across USC’s 23 schools unite to drive innovation. Adam Russell, AI Division Director at ISI, highlights the importance of integration in interdisciplinary work. “It’s not just about combining expertise,” Russell explains. “Effective interdisciplinary work means understanding different perspectives deeply enough to reshape the research questions and solutions.” 

Building these teams starts with mutual understanding and respect. “You need to learn about the other disciplines to have efficient conversations and establish shared use cases,” says Yolanda Gil, Senior Director for Major Strategic AI and Data Science Initiatives at ISI, who has received several awards for bridging AI and Earth Sciences. She carefully identifies collaborators who are open to connecting ideas across fields. “I train my team to be generalists,” she says. “They should look at use cases and see general patterns across different scientific fields.”

“It’s not just about putting experts in the same room,” adds Craig Knoblock, Keston Executive Director of ISI. “It’s about creating a framework where their contributions can truly intersect to address complex problems.” By doing so, ISI tackles challenges in healthcare access, climate science, information security, and more.

Bridging Climate Data Gaps: The LinkedEarth Project

• Collaborating School: USC Dornsife College of Letters, Arts and Sciences, Department of Earth Sciences
• Researchers: Deborah Khider, Research Scientist at ISI and paleoclimatologist; and Yolanda Gil, Director of New Initiatives in AI and Data Science at the USC Viterbi School of Engineering, and Senior Director for Major Strategic AI and Data Science Initiatives at ISI; Julien Emile-Geay, Emile-Geay is a Professor of Earth Sciences at Dornsife

LinkedEarth, co-led by Yolanda Gil and Deborah Khider from ISI and Julien Emile-Geay from the USC Dornsife Department of Earth Sciences, addresses a core challenge in climate science: making diverse paleoclimate data accessible and usable for research. Paleoclimate data comes in many forms, from ice cores to tree rings, but inconsistencies in how datasets are formatted and reported create barriers to analysis. “Each dataset is like its own island,” explains Emile-Geay. “LinkedEarth is about building bridges between these islands so researchers can access and analyze data as a cohesive whole.”

This collaboration began in 2016 with a National Science Foundation grant, following a workshop where Gil and Emile-Geay connected over their shared vision of leveraging AI to standardize paleoclimate data. Emile-Geay provided the project’s foundational questions, including how to align varied datasets to improve reconstructions like the well-known hockey stick graph, which visualizes temperature trends over the past two millennia and is crucial for studying climate change. “Updating this graph with new data takes months,” says Gil. “Julien was eager to speed up that process, and we had semantic technologies that could help scientists in different paleoclimate areas agree to consistent data representations.”

The team standardized terms and reporting methods, creating a framework to represent climate data consistently across sub-disciplines. “This isn’t just about integrating data — it’s about empowering collaboration,” says Khider. “We wanted to make it easier for researchers to share and build on each other’s work.” Through workshops and training sessions, the LinkedEarth team has expanded its reach, empowering researchers to adopt its tools and contribute their own data. 

Early Glaucoma Detection in Sight: The ATLAS Initiative

Collaborators: Keck School of Medicine, USC Dworak-Peck School of Social Work, Los Angeles County Department of Health Services
Researchers: Carl Kesselman, Director of the Informatics Systems Research Division at ISI, the William H. Keck Chair of Engineering in the USC Viterbi School of Engineering; Benjamin Xu, Associate Professor of Ophthalmology at Keck School of Medicine; Shinyi Wu, Associate Professor of Social Work and Industrial & Systems Engineering; Lauren Daskivich, Director of Eye Care Services, LAC DHS.

The AI and Teleophthalmology in Los Angeles (ATLAS) Initiative, led by Carl Kesselman of ISI and Dr. Benjamin Xu of the Keck School of Medicine, tackles a critical global health issue: detecting glaucoma early enough to prevent blindness. Known as the “silent thief of sight,” glaucoma often progresses unnoticed until vision is permanently damaged. Kesselman says, “We have developed new AI-ready datasets and tools that enhanced  the accuracy and reproducibility of our glaucoma detection algorithms.” By using AI to analyze retinal images, the project seeks to detect early signs of optic nerve damage, transforming glaucoma care from reactive treatment to proactive prevention. “Our aim is to create a tool that identifies glaucoma at a stage when intervention can save sight,” Xu explains.

A distinguishing feature of this initiative is its additional partnership with the Los Angeles County Department of Health Services and USC Dworak-Peck School of Social Work, which adds a holistic approach to ensure accessibility. “Early detection isn't just about medical precision — it's about creating a cost-effective solution that brings quality healthcare to underserved communities,” says Shinyi Wu, an Associate Professor of Social Work. “Our approach minimizes long-term healthcare expenses by preventing advanced glaucoma, making our technological intervention both medically and economically impactful.”

The tools developed are adaptable for use in healthcare settings such as primary care offices and community health centers, making them particularly impactful in low-resource environments. “Our goal is to create a solution that works where ophthalmologists may not be readily available,” says Kesselman. In February 2025, the ATLAS Initiative was awarded the NEMO Prize, which supports collaborations between researchers at the USC Viterbi School of Engineering and the university’s constellation of health sciences schools. 

Policies for Progress: The MINERVA Project

• Collaborating School: USC Sol Price School of Public Policy
• Researchers: Andrea Belz, Professor of Practice in the USC Viterbi School of Engineering; Alexandra Graddy-Reed, Associate Professor at Price

The MINERVA Project (Management of INnovation, Entrepreneurial Research, and Venture Analysis) is a collaboration between ISI and the USC Price School of Public Policy that addresses funding trends for technology startups, examining how public and private funding decisions influence the development of transformative technologies. By understanding these factors, MINERVA informs policies that accelerate the transition of technologies from discovery to applications.

“We look at the decision-making processes in both federal programs and private investments, and then we look at the consequences,” explains Andrea Belz, a leader in entrepreneurship and innovation. “We uncover funding patterns that reveal problems on a national scale, and then translate these insights into actionable proposals. Our hope is to move beyond describing problems to formulating potential solutions.”

To achieve this, MINERVA uses advanced data science and machine learning tools to analyze large datasets, such as grant applications and venture capital investments. These tools highlight patterns in the funding landscape and reveal ways to distribute resources for many national needs. “Together, we develop data-driven models and translate them into practical policy recommendations,” says Graddy-Reed. “Our advanced data tools directly inform our proposals to make innovation funding more effective.”

MINERVA emphasizes impact, creating data tools for use by academic researchers and analysis for private investors to support the ecosystem. “At MINERVA, we want to inform the national conversation with rigorous analysis,” says Belz. 

Toward Fairer Justice: Detecting Bias in Legal Language

• Collaborating School: USC Gould School of Law
• Researchers: Jonathan May, Principal Scientist at ISI and Research Associate Professor at USC Viterbi; and Jonathan Choi, Professor of Law at Gould

The Detecting Bias in the Law project, led by Jonathan May at ISI aims at identifying and mitigating biases in legal texts — biases that can shape judicial outcomes and disproportionately affect marginalized groups. The project uses advanced natural language processing models — May’s specialty — to analyze large volumes of text from legal documents such as court opinions, statutes, and policy frameworks, to highlight patterns that may unintentionally reinforce stereotypes or discriminatory outcomes.

“We believe legal language may carry implicit or explicit societal biases that can influence outcomes,” says May. “The goal is not just to detect where biases exist but to understand how they shape legal decisions and work toward fostering fairness.” Jonathan Choi, a Professor of Law at Gould, provides essential legal expertise, identifying areas where biases may occur. “We believe AI can uncover patterns that legal experts might miss,” says Choi. “My role is to ensure those findings are accurate and meaningful within a legal framework.”

The collaboration has also given May a new perspective on data analysis. “As computer scientists, we often focus on data sets without fully considering or knowing their provenance — the origins, transformations, and decisions that shape them,” May explains. “Legal scholars think about data differently. A legal decision isn’t the product of a single source; it’s built from layers of rulings, interpretations, and contexts that can’t be reduced to a single piece of data. I understood that intellectually, but working with Jonathan Choi has been fascinating in seeing how that plays out in practice.”

Quantum Security: Advancing Quantum Communication

• Collaborating School: USC Dornsife College of Letters, Arts and Sciences, Department of Physics and Astronomy
• Researchers: Haley Weinstein, Research Scientist at ISI; Jonathan Habif, Principal Scientist at ISI; Todd Brun, Professor of Electrical and Computer Engineering and Computer Science at Viterbi, and Physics and Astronomy at Dornsife

Haley Weinstein’s work on quantum steganography, a method of hiding communication within what appears to be random noise, demonstrates a unique benefit that can emerge from interdisciplinary collaboration: bringing theory to life. 

This project, conducted with Todd Brun of Viterbi and Dornsife, and Bruno Avritzer, a Ph.D. student in Physics at Dornsife, began with a theoretical protocol devised by Brun and Avritzer. “They developed a framework to make a laser signal indistinguishable from noise,” Weinstein explains. “My advisor, Jonathan Habif, and I realized we could actually bring it to life using currently available technology,” an achievement that is rare in experimental quantum research.

The collaboration bridged two distinct worlds: theoretical physics and experimental engineering. “The theorists saw matrices; I saw voltages,” Weinstein says. “Bridging that gap helped connect the theory to the experiment.” By working together, Brun and Habif's groups proved that the protocol works in real-world conditions. Brun says, “It was remarkable to see the abstract problem that Bruno and I had worked on transformed into something that could actually be done in the lab with realistic technology, and even more exciting to see it produce real data.” Habif adds, “It was inspiring to see two students with very different interests and types of training work closely together, learning from each other along the way. The result was two well-rounded young scientists and an exceptional new technology demonstration in the field of quantum information.” 

Currently, Weinstein is building a full transceiver to send data using the protocol, with potential applications in national security and data privacy. “This collaboration was possible because ISI is uniquely positioned at the intersection of academia and industry,” Weinstein notes. “We have our finger on the pulse of real-world applications while still fostering the academic freedom to explore new ideas.”