AI-Powered Ocean and Earth Science: A Data-Driven Revolution

Unlocking Discoveries with Artificial Intelligence

Artificial intelligence is rapidly transforming scientific research, offering unprecedented insights at a blistering pace. At Oregon State University's College of Earth, Ocean, and Atmospheric Sciences (CEOAS), AI is revolutionizing the study of our planet, enabling researchers to navigate the vast ocean of unstructured data like never before.

This research is fueled by Dell Technologies Inc.'s cutting-edge infrastructure, allowing for real-time analysis of critical environmental shifts. As Christopher Sullivan, director of research and academic computing at CEOAS, explains, "We really do focus on how the planet is being changed by our interactions with it. We use a lot of data, obviously, to help us do that."

High-Performance Data: The Engine of AI-Driven Research

From microscopic plankton to expansive ecosystems, CEOAS research generates mountains of data. Advanced AI models are essential to extracting meaningful insights from this information. Dell's scalable infrastructure plays a pivotal role, providing the necessary computational power to manage this data deluge.

Sullivan highlights the transformation, stating, "Before AI... we were data-rich and information-poor. AI is data-driven, and it's married to that data." This data-driven approach, empowered by Dell's technology, allows researchers to tackle complex scientific questions with unprecedented efficiency.

CEOAS researchers utilize specialized infrastructure tailored to specific scientific inquiries, configuring machines with diverse capabilities to meet varying computational demands. This flexible approach, facilitated by Dell, stands in contrast to traditional enterprise environments that rely on standardized hardware.

"One of the great reasons we use Dell is because we get the ability to build out different platform technologies that really focus in on the questions we’re trying to answer," Sullivan explains. "Different data requires different technologies for processing."

Taming Data Complexity: From Collection to Insight

The massive datasets generated by CEOAS require seamless movement between disciplines. To facilitate this, the university has built a high-performance, on-premises infrastructure optimized for speed and efficiency. Leveraging Dell's advanced computing platforms, CEOAS accelerates AI-powered research, enabling faster insights than ever before.

Sullivan emphasizes the importance of on-premises infrastructure, explaining, "We have too much data to be putting that [work] in the cloud... we need to own that on-prem and move it around between our different disciplines and domains very rapidly." Dell's 800 gig networking solutions have been instrumental in optimizing data interaction and management at the petabyte scale.

Structuring unstructured data, like images and sounds, presents another significant challenge. AI models rely on metadata tagging and annotation to make this information searchable and actionable. Simultaneously, robust governance and security measures are crucial for ensuring appropriate access controls and mitigating cybersecurity risks.

Anthony Dina, director of system engineering at Dell, emphasizes the importance of data structure and security, stating, "There's been a huge demand on providing structure to unstructured elements... We have to think about chain of custody in a whole new way — not just at a file level, but at a data level."

Real-Time Insights: Bringing AI to the Ocean's Edge

Real-time data processing is revolutionizing research at CEOAS. Research vessels, including the RCRV Taani, are equipped with Dell PowerScale and PowerEdge systems and Nvidia GPUs, enabling AI-driven analysis at sea.

This immediate data processing accelerates discoveries and allows for timely interventions. Sullivan highlights the importance of real-time analysis using plankton as an example: "If we use that [plankton] as a monitoring tool, it's really like the canary in the coal mine...We really want to be doing it in real time because if it takes us too long, the data becomes meaningless."

This advanced infrastructure also powers groundbreaking marine ecosystem monitoring. Autonomous robotic vehicles equipped with 8K cameras and AI models analyze billions of plankton in real time, identifying rare species and tracking ocean health with unparalleled precision.

Sullivan explains, "There was no way you could possibly have a human quantify that... Statistically, we're actually making relevance now, and I have a lot more ability to see the rare species."