“Power failures can lead to great economic loss and greatly impact on our daily lives,” said Chen, who was awarded the grant last December. “My goal is to improve people’s lives by ensuring the security and resiliency of our energy system.” 

Electric power systems are a critical component of society that provide power to our homes, businesses, and devices. But when they fail, they can have devastating consequences, said Chen. Disastrous events have previously shut down the electric grid and left millions of people without power, including the 2019 Manhattan blackout and the 2021 Texas power crisis.  

Thanks to advances in technology, many electric power systems now use energy devices that can be controlled remotely through smartphone apps and other Internet-based devices. These devices, known as Internet of Things (IoT)-enabled energy devices, can be found in solar panels, wind generation systems, and electric vehicles, said Chen. They can also be found in commonly used household appliances like air conditioners, water heaters, and electric ovens. 

A Weakness with Potentially Devastating Consequences

The original goal of using IoT-enabled energy devices was to improve operational performance through greater reliability and sustainability, said Chen. However, he said that these devices are weak in one critical area—cybersecurity.  

“IoT-enabled energy devices are easy to hack because they are not built with a high level of security. These devices have limited capabilities, and they are incapable of running sophisticated encryption and authentication mechanisms, which our computers have,” Chen said. “These devices are also often operated under factory settings with a default password, so it can be relatively easy to hack them.” 

Hackers can compromise devices in a coordinated manner, said Chen. The attacker first gains control of a group of IoT-enabled energy devices and then forms an IoT botnet—a network of infected devices that can launch a large-scale attack and disrupt the normal operations of an entire power energy system. 

“This can disrupt the supply-and-demand chain of energy suppliers and consumers. It can also create a power surge that makes our electric grid more unstable and potentially lead to a power failure that causes economic loss and human injury,” Chen said. “The cyberattack initially leads to a local power failure. An energy supplier will try to restore the power, but the power failure could propagate and lead to a major blackout due to the highly complex and dynamic nature of grid operations.”

Increasing Protection in the Field and at Home 

In a two-year-long project, Chen and his team of graduate and undergraduate students will conduct a comprehensive study of modern electric power systems, analyze the behavior of potential hackers, and develop defensive strategies to protect the power systems from cyberattacks. Their overarching goal is to create cost-effective mechanisms to improve the security and resiliency of electric power systems under IoT botnet attacks. Collectively, these mechanisms can serve as a guide for grid operators who are responsible for protecting the electrical power system, said Chen. 

Right now, everyday people can protect their personal IoT devices from cyberattacks by taking one simple step—changing their devices’ default passwords, said Chen. 

“Many people ignore this step and leave their devices in a very vulnerable situation. An attacker can guess their passwords very easily and have complete control over their devices,” Chen said. “We also need to regularly patch and update the software systems on our devices, just like we do with our smartphones.” 

Chen said that his team’s research results will be integrated into a new course at Fordham called Artificial Intelligence for Cybersecurity. The course, which will provide students with cross-disciplinary training in cybersecurity, artificial intelligence, and informatics, will potentially be offered in 2023. 

“What excites me most is the nature of this project,” said Chen. “This is a societal problem that will potentially have a lot of impact on our daily lives.”

The 2021 Tri-State ExploreCSR workshop, which began in February and wrapped up last month, aims to fix that.

The workshop brought together students and faculty from Fordham, the Stevens Institute of Technology, the University of Connecticut, and Google to learn or brush up on coding and computer science-related research. On April 23, the group celebrated with online presentations, a panel discussion, and poster board presentations on team projects.

Funding and Participation from Google

Ying Mao, Ph.D., an assistant professor of computer science at Fordham, said the project, which was funded by an $18,000 grant from Google, was created to break down barriers that students from all underrepresented populations might feel are keeping them from entering computer science.

“When we wrote our proposal, we actually broadened the domain of who we wanted to offer it to; usually it’s just for women. We included first-generation students who might not have imagined doing something in computer science,” he said.

The April event not only brought the students together, it allowed them to hear from experts in the field. It started off with a keynote address, Combinatorial Fusion Analysis: A new paradigm for combining multiple scoring systems, which was delivered by Frank Hsu, Ph.D., the Clavius Distinguished Professor of Computer Science at Fordham. He was followed by Adam Michael Wood of Google, who talked about Tensorflow, an open-source software platform that can be used to create machine learning models.

A panel discussion on the challenges and opportunities in computer science featured Fordham’s computer science professors Gary M. Weiss, Ph.D. and Damian Lyons, Ph.D., and the Stevens Institute’s Ye Yang, Ph.D., and Grace Bae, Ph.D.

Organizers put out the call for the in the beginning of the year and received 220 applicants from undergraduate students enrolled at 16 colleges throughout the tri-state region. Some were computer science majors looking for research opportunities; others were non-computer science majors. After an initial meeting on Feb. 22, the 58 chosen students were divided into 18 groups that met weekly to report on their progress.

Instilling Confidence

“Some of the research is very introductory level because the students are not from computer science backgrounds. One of my students was an economics major, so we designed a beginning research project so they could learn what computer science research looks like, to help them build self-confidence,” Mao said.

“They all showed how they learned to code and analyze the data, and since every group contained at least three students, how learned how to communicate with each other to build the project.”

Collaborating with Students from Other Universities

One of the students who attended was Navpreet Kaur, a junior at Fordham College at Lincoln Center. A math/economics interdisciplinary major, Kaur got involved in the field in earnest this fall, when she took her first class as part of a computer science minor. Under the guidance of Fordham’s Juntao Chen, Ph.D., the Ozone Park, Queens, resident worked with fellow Fordham student Maria Jara, as well as a student from Carnegie Mellon University and one from Hofstra University. Their project was titled “Fairness and Privacy in the Optimal Transport for Resource Allocation.”

The project involved applying a mathematical model used to determine how to allocate resources such as workers to places such as factories and mines so that they’re allocated fairly as well as efficiently. It’s the kind of formula that could also be applied to the distribution of vaccines.

“It seemed like a good opportunity to get experience, which I thought would help with internships or grad school, and just understanding how comp/sci really works,” she said.

Participating in the program introduced her to Fordham’s STEM program, and Kaur said she wants to take Dr. Chen up on the offer for more research opportunities next year. Working on the project made her feel confident that she’d be ready to participate in a larger research project that otherwise might seem overwhelming.

“I’m also glad we did it virtually since we were able to work with students from different universities with different backgrounds and different years,” she said. Working together with people with different experiences was very helpful.”

Mao said he plans to apply again for the grant next year and hold another workshop at a date to be determined.