Reimagining Research Infrastructure: AI, Data Sovereignty, and the Future of Health Tech
In the rapidly evolving intersection of biomedical informatics, proteomics, and artificial intelligence, Dr. Jim Green stands at the forefront. His work explores some of the most promising frontiers of science—from predicting protein interactions to designing assistive technologies for aging populations. We talked with Dr. Green, professor in the Department of Systems and Computer Engineering at Carleton University, to understand how high-performance computing (HPC) shapes his research and what Canada can do to stay ahead in the AI-driven biomedical revolution.
High-Performance Demands, Human-Centric Goals
Dr. Green’s lab relies on HPC across all their projects. Whether developing deep learning models for analyzing streaming video or predicting 250 million protein interactions in the human proteome, the scale of the data and complexity of the models are staggering. "We’re in a constant battle to keep our compute at the state of the art," he explains. The primary bottleneck? GPUs—powerful, memory-rich ones that can support cutting-edge AI architectures.
The challenge is not just technical—it’s pedagogical. Students experimenting with new ideas often face long queue times on shared infrastructure, disrupting their creative and iterative workflow. That’s why Dr. Green prioritizes giving every graduate student at least one dedicated GPU. "Each student requires thousands of dollars in hardware to properly develop their research," he notes.
Research That Protects, Learns, and Adapts
One of the lab’s recent focuses is on AI-powered assistive technologies to support healthy aging. Working with Professor Rafik Goubran, Jim is developing privacy-preserving solutions like thermal cameras and pressure-sensitive mats that can monitor respiration and movement non-invasively. "Older adults want to live independently. If we can tell that a loved one hasn’t left bed all day, that should trigger a wellness check," he says. Such technology blends humanity with machine learning—an approach that exemplifies Computing for Humanity’s values.
Unlike wearables, which require user compliance, these environmental sensors function quietly and continuously, reducing gaps in monitoring. “Compliance is a big issue. If someone forgets to wear their device, we lose all data,” Dr. Green points out.
The Need for Sovereign, Specialized Infrastructure
Some of Dr. Green’s work involves sensitive healthcare data, like video from hospital rooms. That kind of data cannot leave the campus, let alone be placed on shared or cloud-based resources. “We need on-premise computing and storage,” Dr. Green emphasizes, especially given growing concerns about data privacy and the sovereignty of Canadian research data. “Wherever your cloud is hosted, do they share our values? Are you at risk of being collateral damage in an attack?”
He sees a national AI platform built within Canada as vital—offering sovereignty, specialized infrastructure, and localized training that’s tailored to Canadian researchers' needs.
Sustainable Innovation and Education
While environmental sustainability hasn’t been top-of-mind for Dr. Green’s team, he acknowledges the role electricity sources play. “Ontario’s hydro means a lower carbon footprint. But if your cloud is in a jurisdiction running on coal, the environmental cost is higher.”
When it comes to preparing the next generation of researchers, Dr. Green sees two parallel tracks: one for computational scientists pushing AI forward, and another for application experts—biologists, clinicians, and other domain experts—who benefit from using AI tools. “Training both groups is essential. AI can help summarize thousands of papers or brainstorm new research directions. That’s powerful.”
Building Bridges: Researchers, Infrastructure, and Charities
For Dr. Green, interdisciplinary collaboration is where the magic happens. "Clinicians, researchers, machine learning experts—we all need to work together. Compute providers should be part of that dialogue. And yes, charities too."
He sees potential in organizations like Computing for Humanity acting as vital bridges—connecting researchers in need with donors excited to support innovation. “A lot of people would love to contribute to science but don’t know how. Connecting them directly to those doing the work makes perfect sense.”
Beyond the technical sophistication, what drives Jim Green’s research is a deep commitment to human well-being. His team’s work on assistive technologies—like thermal cameras for non-invasive respiration monitoring or pressure-sensitive mats to ensure wellness checks—speaks to a powerful intersection of compassion and innovation. These tools aren’t just about data; they’re about dignity, particularly for elderly individuals who wish to live independently and safely in familiar environments. By designing systems that respect privacy while delivering critical health insights, Green’s lab exemplifies how AI can be deployed ethically to improve lives.
Looking ahead, Green envisions a research landscape where access to cutting-edge technology is democratized, and AI training becomes standard across disciplines. He’s passionate about equipping not only computational scientists but also clinicians and biologists with the tools and understanding to harness AI responsibly. But that potential can only be fully realized when Canada invests not just in infrastructure, but in people—students, researchers, and practitioners—who can lead the next generation of discovery.