U.S. Coast Guard accredits analytical system developed at Purdue
May 7, 2013
In efforts to prioritize and efficiently manage the repair of boats and stations damaged by Superstorm Sandy, the U.S. Coast Guard has accredited a system called Coast Guard Search and Rescue Visual Analytics (cgSARVA) developed in collaboration with Purdue University.
The Coast Guard accredited the system on April 22, 2013, at its headquarters in Washington, D.C.
The cgSARVA tool was created by researchers at the Purdue-led center Visual Analytics for Command, Control and Interoperability Environments, or VACCINE, a U.S. Department of Homeland Security Center of Excellence.
"The accreditation is the first time anything produced by a DHS Center of Excellence has been verified and validated for use by the Coast Guard," says David Ebert, VACCINE director and Silicon Valley Professor of Electrical and Computer Engineering. "The cgSARVA tool can help DHS agencies and law enforcement agencies across the country."
The tool has enabled an interactive visualization, analysis and assessment of search-and-rescue missions completed by each Coast Guard station in hurricane stricken parts of New York and New Jersey.
"The cgSARVA tool is especially helpful in guiding operations and resource decisions by carefully analyzing data in a way that ensures the best return on investment," says Vice Adm. Rob Parker, Coast Guard Atlantic Area commander. "This project serves as a great example of positive partnerships that are being forged between the Coast Guard, the DHS Center of Excellence, and academia."
Purdue initially designed the computer-based visualization to help Coast Guard analysts assess adjustments to boat stations and capabilities on the Great Lakes. It was later used in the Mid-Atlantic region to reallocate resources for Hurricane Irene in 2011 and last year in the aftermath of Superstorm Sandy, which severely damaged 14 Coast Guard stations in the region. The Coast Guard is using the tool to prioritize rebuilding of damaged stations and to help determine which stations should and shouldn't be rebuilt.
"The system can look at what happens if you were not able to immediately rebuild a given station with a certain search-and-rescue caseload," Ebert says. "How long would it take other stations to respond if this station were not here? And if this station were not here, how many cases would have to be handled simultaneously by nearby stations? So it doesn't take all the input and give a final answer, but it provides criteria of the workload and the benefit and what happens if a station closes."
Following Superstorm Sandy, Coast Guard analysts were charged with prioritizing the rebuilding of damaged small-boat stations to determine the order in which stations were to be repaired.
"The cgSARVA model formulation proved to be tremendously insightful for the Coast Guard as it began to prioritize the repair of its stations," says Commander Kevin Hanson, analysis team leader. "Even upon receiving full funding for all damages, the Coast Guard is unable to execute all repairs at the same time and the outputs from cgSARVA have been instrumental in assisting senior leadership in prioritizing work."
Using cgSARVA, the Coast Guard was able to quickly and easily determine how resources might be reallocated in New Jersey, allowing the Coast Guard to operate with increased efficiency.
"A remarkable amount of intellectual rigor has brought us to this point," says Rear Admiral Dean Lee, Deputy for Operations Policy and Capability. "Our analysis team here at headquarters saw tremendous potential in the initial version of cgSARVA and had the organizational vision to expand its capabilities for inclusion in their strategic modeling efforts. Our partnering with Purdue University and the Research and Development Center has yielded insight into our coastal operations that we have never achieved before."
Three Purdue graduate students have been involved in the cgSARVA project, which is ongoing, with researchers continuing to add new capabilities.
The computer-based modeling tool runs on an ordinary computer or laptop.