Detecting threats and emergencies to protect infrastructures

KALAMAZOO, Mich.—In a relatively new type of research, Dr. Shameek Bhattacharjee, assistant professor of computer science, envisions a time when emergencies and cyber threats in infrastructures such as city transportation systems or electricity service distribution systems are detected and reported to responders almost immediately, resulting in better service and lower costs. 

Through his research, Bhattacharjee collaborated with researchers from Missouri University of Science and Technology to develop this novel, unified anomaly detection framework that can detect such emergencies and provide an early indication of cyber-attacks, which applies to a variety of smart technologies for civic infrastructure, including smart transportation and smart electric grids.

“Our framework can be applied to various systems that need vulnerability detection and has multiple stages,” says Bhattacharjee. “In the first stage, instead of trying to pinpoint whether the anomaly is due to an emergency or an attack, we simply detect that there is a problem quickly and alert responders who can initiate appropriate workarounds until the problem is corrected. In the second stage, our framework tries to determine the specific components that have been affected and isolate them from the rest of the system. Because our framework is based on simple operations, it doesn’t require expensive devices or a lot of computing power.”

Clearing the way in Nashville

Many people dread driving through a city. Accidents, construction projects, protests, celebrations, heavy traffic, blinding rain, and aging infrastructure can make the trip a nightmare. To alleviate the time required to clear your path and also to aid quick responder dispatch, Bhattacharjee is applying his framework in Nashville on a smart transportation project using data collected from roadside sensors.

“This work is a good example of a public-private partnership,” says Bhattacharjee. “Combining these two pieces, we proposed a new method using our framework to automatically detect incidents that cause traffic disruptions and pinpoint the exact location of the incidents in less than five minutes. This enables quick and transparent re-routing of traffic and also may be used by responders and emergency services.”

Extension to the electric grid

As more green sources of energy are developed, changes to the electric grid system are becoming essential. Because these changes require connecting the grid to the internet for operations, security threats increase, adding complexity to the transition. Bhattacharjee and Missouri S&T’s Dr. Sajal Das have begun applying the framework to security vulnerabilities in electricity distribution systems in smart electric grids. Their research has been published in leading journals from Institute of Electrical and Electronics Engineers and the Association for Computing Machinery.

“Detecting these anomalies, or threats, is inherently far-reaching and critical,” he adds. For Bhattacharjee, his work both challenges his research acumen and enhances his life’s meaning. “My work related to smart grids is associated with a social cause that gives me a sense of purpose.”

His research is supported by National Science Foundation.