iRODS data system powers cutting-edge scientific collaboration

by Paul Tooby, DICE Communications

Each weekday, about a quarter of the U.S. population is engaged in the critical business of educating tomorrow’s society - including 56 million students in K-12 schools, 18 million in college, plus all of their teachers.

But this massive enterprise is facing steep challenges, and as the nation seeks solutions for its troubled educational system, the NSF Temporal Dynamics of Learning Center (TDLC) at UC San Diego is bringing together scientists from many disciplines to deliver accurate scientific knowledge that educators can use in their classrooms.

"Understanding how people learn is far too complex for any one discipline, so we’re building on the synergy of scientists with different expertise," said Andrea Chiba. "This brings together the benefits of diverse approaches from neuroscience, cognitive science, computer science, physics, and psychology."

But it’s no easy task to cooperate with more than 50 researchers across a dozen institutions doing cutting-edge work in a range of disciplines. To weave these distinct strands of scientific work into a coherent fabric of knowledge, the scientists need to manage the disparate types of information they gather.

A unifying theme of the research is the role of time in learning. For example, in studying social interactions in learning, the researchers ask: How does the brain organize data as a student is in the process of learning something new? To capture the fleeting steps of this complex process means gathering data in real time, using 256 head-mounted sensors which gather brainwave data. At the same time the scientists gather simultaneous audio and video data from 12 cameras.

"It’s a data management nightmare," said Chiba. "Gathering data is the easiest part, we then need to organize, describe, and make sure the data is easy to share among a number of collaborating groups for analysis." For example, one research group uses computers to extract information from the video to study facial expressions while another group uses the brain information to understand how it correlates with learning.

Because the data involves human subjects, the scientists have to obtain Institutional Research Board approval for distribution of the data and implement stringent access and security controls.

The data system the researchers found that can best address all their data needs, from capturing and storing massive amounts of data to applying human subjects protocols and sharing across multiple institutions, is the new NSF-supported Integrated Rule-Oriented Data System (iRODS) developed by the Data-Intensive Computing Environments group at the School of Information and Library Science at the University of North Carolina at Chapel Hill and the UCSD Institute for Neural Computation at the University of California, San Diego.

For the scientists and NSF, an important bonus is that the iRODS system can also automate complex administrative and information reporting tasks, freeing their time to focus on research on the exciting science of learning.

The iRODS software is open source, benefiting from broad collaborative development by international groups. This leverages the investment by the National Archives and Records Administration (NARA), the National Science Foundation, and other agencies and ensures that the benefits of the technology are sustainable and reach scientists across a wide range of disciplines.