North Carolina State University, a leading US public research university, is home to more than 34,000 undergraduate and graduate students who learn by doing. A powerhouse in science, technology, engineering and math, NCSU supports ground-breaking research across a range of disciplines.
The Center for Geospatial Analytics at NCSU is the foremost interdisciplinary research and teaching center of its kind in the United States that is discovering, interpreting and communicating meaningful patterns in location-based data, which can include satellite images, temperature records, soil moisture levels, and even social media posts.
The center sifts through this data to find patterns that help them better understand everything from large-scale climate events like floods, droughts, and wildfires, to crop yields.
That’s a lot of data. And a lot of data is always a challenge.
Fighting Data Deluge
Around 70% of data collected worldwide is spatial and temporal data. This means that the data contains both X and Y coordinates, such as latitude and longitude, and a timestamp of when the observation took place.
“These spatiotemporal datasets that we need to analyze are very large and heterogeneous. Traditional machine learning algorithms cannot deal with these high dimensional heterogeneous datasets. AI and deep learning technologies are perfectly suited to this type of work, but demand high computational resources and expertise that are in shortage,” says Dr. Raju Vatsavai, Associate Director of Spatial Computing and Technology at NCSU.
Another problem that most research institutes face is funding. “Due to academic funding limitations, it’s becoming harder and harder for centers like ours to buy high-performance computers with the latest technology,” says Vatsavai.
In its attempt to solve some of the world’s biggest challenges, NCSU knew that in order to succeed it needed access to high-performance computing and AI expertise. And it realized it didn’t have to look too far.
AI Lends a Hand
With help from Lenovo, the Center for Geospatial Analytics is taking advantage of AI to tackle some of the biggest challenges of our age.
Vatsavai and his team now have access to the newly available Lenovo Artificial Intelligence Innovation Center, located in Morrisville, NC. The Lenovo AI Innovation Center is kitted out with a wide array of high- performance computing resources, including the latest Lenovo ThinkSystem servers which are equipped with Intel® Xeon® Scalable processors offering next-level performance.
“Being able to access such powerful computing resources optimized for AI is fantastic, and it means that researchers can run more data analyses faster, accelerating projects,” says
Dr. Raju Vatsavai. “Having access not just to the Lenovo technology, but also to Lenovo’s expertise, is extremely valuable. They helped us to develop new algorithms that will enable us to analyze many more data streams in near real time.”
This will help NCSU find new patterns that will get us closer to understanding and solving some of humanity’s biggest challenges.
“Our hope is that with new AI capabilities, we can analyze sensor data from farms across the US and, based on regional weather patterns, predict the best time to irrigate crops to minimize wastage and misutilization of water supplies, boosting efficiency. Behind all the number-crunching, it’s about changing the world.”
Share This Article