January 15, 2014

NEES-led data integration project aimed at exposing U.S. quake data to researchers in Europe, elsewhere globally

Martinez and Marshall

Ignacio Lamata Martinez, left, a researcher at the University of Oxford in England, and Gemez Marshall, senior software engineer for NEEScomm IT at Purdue, discuss efforts of the team at the George E. Brown Jr. Network for Earthquake Engineering Simulation to develop a system for integration of shared experimental data for global use. (Purdue University photo/Mark Simons).
Download Photo

WEST LAFAYETTE, Ind. - Researchers and software developers at the Purdue-led George E. Brown Jr. Network for Earthquake Engineering Simulation (NEES) are building a data integration system that will allow seamless sharing of experimental data with earthquake engineers globally.

The innovative system, which is using Semantic Web technologies, will initially be made available to earthquake engineering researchers in Europe before opening up to investigators in Asia, South America and elsewhere.

U.S. earthquake engineers from Buffalo to Berkeley have been conducting research sponsored by the National Science Foundation aimed at improving structural seismic design and reducing the damaging effects of earthquakes and tsunamis. Making that research usable to researchers across the globe, however, is a challenge because there are no common international standards for describing and organizing multihazard engineering data.

A team at the NEES Operations headquarters (NEEScomm) in Purdue's Discovery Park, in seeking to encourage reuse of the 20 terabytes of experimental data in the NEES Project Warehouse, is leading a collaborative effort to provide a system of standards necessary for meaningful international data exchange in the field of earthquake engineering.

The new system, called Celestina Data, will enable data interoperability without requiring researchers to modify their current working models. Instead, Celestina will provide an ontology-based data access layer that exposes data - and makes it usable - to researchers anywhere.

"We are aggressively working to provide practical technological solutions so that researchers who examine how to minimize the devastating effects of earthquakes and tsunamis can communicate and more efficiently find answers," said NEES director Julio Ramirez, a civil engineering professor at Purdue. "It's not a simple task, but this achievement will go a long way toward strengthening the earthquake engineering community across the globe."

Joining Ramirez on the Purdue team are mechanical and civil engineering professor Shirley Dyke along with NEEScomm IT director Brian Rohler, senior software engineer Gemez Marshall and data curator Standa Pejša. Oxford University researcher Ignacio Lamata Martinez, also on the Purdue team, serves as lead developer of Celestina.

Lamata Martinez, who has built distributed systems for the Seismic Engineering Research Infrastructure for European Synergies (SERIES), is driving this collaborative data-interoperability effort at Purdue for engineering researchers around the world. Lamata Martinez outlined details of the data integration system for earthquake engineering in a co-authored research paper published in the Journal of Computing in Civil Engineering in April 2013.

In addition to collaborating with Oxford University in England, the Purdue team has formed key institutional partnerships with the European Commission's Joint Research Centre in Ispra, Italy.

The goal is to develop the system prototype, which will be fed by relevant databases in the U.S. and Europe, with a tentative rollout by September. At that point, the group will engage international partners to expand this data integration effort. Researchers and students interested in seamless access to earthquake engineering data are in Japan, China, South Korea, and elsewhere -"anywhere in the world that deals with earthquakes," Dyke said.

Dyke said such access to even greater amounts of data will accelerate the dissemination and use of research results across Europe, while moving the latest earthquake engineering research into practice, driving innovations and enhancing risk mitigation.

Through this approach, research data having an impact on design decisions, construction materials, building codes and geotechnical issues would continue to be housed at regional databases, perhaps from a specific nation or individual laboratory.

But these regional databases, once "plugged into Celestina's common ontology layer," will be seamlessly accessible and usable to engineering researchers around the world, she said.

"Enabling shared data and tools from any access point around the globe means that NEEShub users will be able to view, download and use remote data from other connected databases directly through the NEEShub," Dyke said. "Currently, the fact that there are numerous isolated sources of such data creates barriers that are difficult and expensive to overcome for researchers who develop the innovations necessary to minimize damage during earthquakes and tsunamis."

Marshall said the team's primary goal is to facilitate seamless collaboration within the earthquake engineering community, linking the network of researchers across the globe and applying Semantic Web technologies as a means to further research and knowledge.

"As you might guess, there are technical and social challenges to the success of this idea. Our mission is to make it happen," Marshall said.

In 2009 earthquake engineers and researchers in Europe formed SERIES to address earthquake-engineering needs, including data sharing for research generated at eight major sites and 10 shake table laboratories across the continent. More than 20 universities and research institutions have been a part of SERIES.

"SERIES has developed a virtual database that has greatly enhanced the impact of European research infrastructures by improving their capacity for data exchange, sharing and access," Lamata Martinez said. "It has achieved many noteworthy successes, including developing a common portal for access to the distributed repositories at the laboratories, as well employing telepresence and advancing distributed testing."

The Purdue project also centers on Project Warehouse, the NEES data repository that was created through a National Science Foundation effort to enhance earthquake-engineering capabilities in the U.S. by creating a network of shared laboratory facilities and archiving and publishing resulting research data.

Based at Purdue's Discovery Park, NEES is a collaborative, 14-site research initiative that aims to improve structural seismic design and reduce the damaging effects of earthquakes and tsunamis. NEES is funded by a $105 million NSF grant.

Since Oct. 1, 2009, the NEES operations and cyberinfrastructure headquarters has been located at Purdue's Discovery Park, the result of National Science Foundation cooperative agreement #CMMI-0927178. NEEScomm is the operations unit at Purdue. 

Researchers anywhere in the world can use the network, view real-time experiments and even run interactive simulations. The 14 sites are at major research universities and the network is managed at Purdue, which provides the IT research and cybernetwork, powered by HUBzero software developed at Purdue. 

Writer: Phillip Fiorini, 765-496-3133, pfiorini@purdue.edu

Sources: Julio Ramirez, 765-494-2716, ramirez@purdue.edu

Shirley Dyke, 765-494-7434, sdyke@purdue.edu

Ignacio Lamata Martinez, 44-0-1865-2-83469, ignacio.lamata@eng.ox.ac.uk

Gemez Marshall, 765-496-6445, gemezm@purdue.edu