Search form

Advanced Search
 
Monday April 24, 2017

Detecting Nuclear Threats, Rogue Ships and Bad Actors

Wednesday April 5, 2017

Detecting Nuclear Threats, Rogue Ships and Bad Actors

Your Source for University News
Rutgers professors use statistics to try to prevent terrorist attacks and solve societal problems

Photo: ArtOlympic/Shutterstock
Some police cars have sensors to detect nuclear and other radioactive material.
Could statistics help detect radioactive “dirty bombs” in big cities, drug-smuggling ships on the high seas and other threats?

The odds are excellent, two Rutgers professors say.

Min-ge Xie and Rong Chen, faculty in the Department of Statistics and Biostatistics in the School of Arts and Sciences, helped devise statistical algorithms and models for enhancing the detection of nuclear and other radioactive material, such as in dynamite-based dirty bombs, in cities. Their statistical tools could also be used to uncover other threats, security problems and criminal activities.

“We can estimate the probability of real threats versus false alarms and the police can decide how to respond,” said Chen, a distinguished professor who directs the Financial Statistics & Risk Management Program and the Data Science Program. “The goal is to pinpoint a cluster of hits, strong evidence of a real threat.”

The Rutgers statisticians – funded by a National Science Foundation (NSF) grant along with Cliff Joslyn from the Pacific Northwest National Laboratory – came up with statistical models for detecting nuclear and other radiation as part of the foundation’s Algorithms for Threat Detection program.

Xie and Chen also worked with Fred S. Roberts, a distinguished mathematics professor who directs the Command, Control, and Interoperability Center for Advanced Data Analysis at Rutgers, on a similar project jointly funded by the NSF and the federal Domestic Nuclear Detection Office.

Statistical tools showed that placing radiation sensors and GPS tracking devices in many vehicles with random routes, such as taxicabs, could greatly increase detection of a nuclear threat in urban areas.

The tools work by determining whether a significant radiation cluster exists in an area or if it’s just a false alarm. Once a location is pinpointed, law enforcement could be dispatched to check it out. Some police cars in New York and other cities have radiation sensors but usage is still limited, and there are not nearly as many police cars as taxis, meaning the probability of detection is not as high.

The same concept can be applied elsewhere. In war zones, for example, soldiers could carry chemical detectors to quickly detect chemical weapons, said Xie, a distinguished professor who directs the Office of Statistical Consulting. Citizens with cell phones could text about suspicious activities, and many reports near a location would reveal something that should be checked out.

Today, the professors’ goal is to develop statistical tools to enhance maritime safety and detect real-time threats to national and global security from shipping, including human and drug trafficking, smuggling, transport of nuclear material and dirty bombs, garbage dumping and illegal fishing.

“These tools could also help track down threats to cybersecurity and money laundering,” said Chen, an expert in assessing vast amounts of data.

“With statistics, you’re trying to help solve real-world problems,” said Xie, a pioneer in “fusion learning,” which combines information from different sources to make more accurate and more efficient estimates and predictions. “You’re trying to really help make a difference.”


For media inquiries, please contact science communicator Todd B. Bates at tbates@ucm.rutgers.edu or 848-932-0550.

Your Source for University News