Semper Paratus, the motto of the U.S. Coast Guard, means “Always Ready.” This rings true in the life of USCG Commander Blair Sweigart M.S. ’14, Ph.D. ’19, whose work as a graduate student at William & Mary and as a service member may help the United States better prepare for national security threats.

Working with Professor of Mathematics Rex Kincaid, Sweigart used his knowledge of Network Location Theory to advance ways to track criminal, terrorist or dark networks. His work, on which he centered his dissertation, was recently featured in the peer-reviewed journal Military Operations Research.

“The resource scarcity that is prevalent throughout law enforcement often dictates new approaches to problem sets to maximize efficiency,” said Sweigart, chief of modeling, simulation and analytics with the USCG Research and Development Center. “Operations research has long been used to optimize resource placement and usage, and this work leverages those tools towards the fight against terrorism and criminal networks.” 

From the water to the classroom

Sweigart’s almost 20-year USCG career includes assignments fulfilling missions in port and waterway security, search and rescue, law enforcement and marine environmental protection.

USCG Commander Blair Sweigart M.S. ’14, Ph.D. ’19 (Courtesy photo)

Coming off his third tour, Sweigart – who holds a Bachelor of Science from the USCG Academy — decided to pursue a master’s degree in computer science (computational operations research) at William & Mary, where his wife had graduated.

While a full-time master’s student, Sweigart overloaded his coursework, positioning himself to also complete a doctorate in applied science. After his qualifying examinations, he became USCG chief of maritime law enforcement in Seattle and worked long days responding to emergency situations while also completing his dissertation with Kincaid serving as an academic advisor and dissertation committee chair.

From theory to application

Sweigart’s interest in Network Location Theory emerged from a directed-research summer experience with Kincaid. NLT is an area of mathematics concerned with the placement of objects on a network or grid or some underlying structure modeled as a network or a grid. Sweigart focused on integer linear programming formulations or models in the closely-related areas of Open Locating-Dominating Sets, Locating-Dominating Sets and Identifying Codes, working to advance understanding of optimal sensor placement to detect and immediately locate an event.

These three areas generally examine where best to place sensors within a network, or grid, so that whenever an event occurs two conditions are met. The first condition is that for anywhere in the network, at least one sensor is close enough to detect the event. The second stipulates that every location in the grid is covered by a unique set of sensors, so that the event’s location is immediately apparent based on which sensors report.

In developing generalized models, Sweigart used computer applications to test the formulations on datasets ranging from five to 1,000 nodes to validate the algorithm accuracy and monitor computational demands. Sweigart then used the formulations to examine two specific use cases presented in the research literature pertaining to the terrorist networks behind the U.S. attacks on Sept. 11, 2001, and the Paris attacks in 2015.

Rex Kincaid, Chancellor Professor of Mathematics at William & Mary, served as Sweitgart’s academic advisor and dissertation committee chair. (Photo by Stephen Salpukas)

Sweigart’s unique contribution was the Mixed-Weight Combined Locating-Dominating formulation, which both extended the detection radius to detect activity at greater distances and examined how the formulations could represent the potential effectiveness of different monitoring techniques, including surveillance targets and confidential informants deployed with various levels of intensity.  He also developed a final model that combined multiple formulations into a single mixed-weight construct to identify optimal sensor placement and sensor type.

Sweigart presented his preliminary findings at conferences and obtained vital feedback from pioneers in NLT. He collaborated with USCG and other military personnel, the FBI and other law enforcement agencies interested in potential application of these.

“Sweigart’s work made heavy use of modeling, and he successfully connected with people in the military and law enforcement fields to apply the models to a security and terrorism context,” said Kincaid.

Sweigart’s novel contributions can lead to the broad use of these mathematical constructs to combat foreign and domestic terrorists where the actors may be already known and in dark networks where there is little knowledge of the perpetrators, he said. Refined through additional research, his applications could become a valuable complement to current best practices.  

An additional promising area of research on this topic, says Sweigart “would be to incorporate a temporal or time component that could monitor event impact spread through a network, tracing back to an original source.” This, he thinks could accelerate the adoption of the use of these formulations in the fight against terrorism.