Four University of Rhode Island engineering professors have been awarded a $4.7 million grant from the Office of Naval Research to advance the security and resilience of AI-enabled power grids, promote workforce development, and secure manufacturing environments.
The three-year project, “Advancing Research on Cyber-Physical Security and Resilience: A Multifaceted Approach,” will be led by principal investigator Yan (Lindsay) Sun, professor and department chair of Electrical, Computer and Biomedical Engineering, in collaboration with Hui Lin, Kaushallya (Kay) Adhikari and Manbir Sodhi, and will be managed by the URI Center of Cyber-Physical Intelligence and Security Center.
Cyber-physical systems are at the forefront of modern technology, representing a significant integration of computation, communication, control, and physical processes. Examples of CPSs include power systems and manufacturing, along with transportation networks and health-care infrastructure, highlighting the diverse applicability and critical nature of these systems.
The workforce development components in the project include assistance for two online graduate certificate programs set to launch in Spring 2025. The Future Autonomous Systems Certificate and the Industrial 4.0 Certificate, designed by distinguished professors, are tailored to meet the needs of today’s diverse workforce. These programs aim to equip professionals with cutting-edge skills in critical engineering systems, preparing them to excel in an increasingly complex technological landscape.
Critical infrastructures at risk
Cyber-physical systems, like power grids and manufacturing systems, are critical infrastructures. Security and resilience are central to CPSs due to the risks posed by the interconnectedness of the cyber and physical worlds, the increased level of automation, and the involvement of artificial intelligence. Protecting these systems against adversarial environments is crucial. It involves a multi-layer approach to defense that includes preventive measures, anomaly detection and mitigation, and self-healing.
Consequently, any disruption in critical infrastructures can cause failure in other applications relying on them. “The disruption can lead to unrecoverable damage, e.g., psychological phobia, economic losses, and human casualties. If the disruptions affect military infrastructures, this may further lead to the threat of national security,” said Assistant Professor Hui Lin.
Sun has been researching power grid security for more than a decade, and her work has been applied to cyber-physical security of various engineering systems. She founded and heads up the URI Center of Cyber-Physical Intelligence and Security. “The better we understand how the power grid behaves, the better we can protect it,” added Sun. “There’s the big picture aspect, but under that umbrella there’s a lot of innovation in there.”
Increasing resiliency is essential
Resilience relates to a system’s ability to maintain essential operations during adverse events and extreme operating conditions. There need to be methods in place to continue essential functions under less-than-ideal circumstances.
Manbir Sodhi, professor of mechanical, industrial and systems engineering, has extensive experience with defense and commercial organization systems. He has specialized in manufacturing optimization, supply chain modeling, the Internet-of-Things, and manufacturing tracking.
“Manufacturing systems have become complex machines with hardware and software (and people) all working together. They are susceptible to the same concerns of security vulnerabilities, complexity and reliability,” said Sodhi, noting his research has investigated these facets of systems operations in both theoretical and applied projects. “In this work, I plan to use the modeling and analysis tools for systems analyses and optimization for defense and commercial manufacturing systems, and to develop theoretical advances as well as a test bed at URI for demonstrating our progress in advances related to power systems and manufacturing systems research.”
Lin’s expertise in intrusion detection systems will aid in programming the network to the cyber-physical test bed. “The biggest challenge is that interdisciplinary research requires close collaborations between researchers from different areas. However, many research disciplines are still very close ecosystems, making understanding the terminology in that domain difficult,” said Lin.
Another challenge to this type of project is the lack of visibility of this research domain to public audiences. “I hope that many people can understand it reasonably, realizing its importance but not fearing the over-exaggerated facts,” said Lin.
Crucial steps forward
Assistant Professor Kaushallya (Kay) Adhikari has demonstrated expertise in sensor array design and statistical analysis, particularly in the field of detection and estimation theory. In this project, she will be estimating and predicting short-term power generation.
“Our project focuses on advancing our understanding of solar energy generation, a crucial step toward transitioning to renewable energy sources and reducing our carbon footprint,” said Adhikari. Traditional energy sources release significant amounts of carbon dioxide, contributing to climate change and environmental degradation. By studying and understanding renewable energy generation, the aim is to mitigate these harmful effects and pave the way for a cleaner and more sustainable future. “Through our efforts, we hope to accelerate the adoption of renewable energy technologies and contribute to global efforts to combat climate change.”
Her previous research involved developing sensor arrays for various applications and analyzing the data collected from the arrays. Her experience in estimation theory will be invaluable for analyzing renewable energy data. “While my previous work primarily focused on acoustic, electromagnetic, and EEG data, I am eager to apply my knowledge to the domain of energy generation,” said Adhikari.
Preparing the next generation
By accounting for as many relevant factors as possible in energy level predictions, the team aims to provide accurate and reliable estimations that can inform decision-making processes at scale. They can then test estimation methods on simulated data, refine their models, and validate their performance under controlled conditions. Once they have established effectiveness in simulation, the testing transitions to theories on real-world data, and ensures findings are robust and applicable in practical scenarios.
Each of these areas is critical for ensuring the robustness of cyber-physical infrastructure and preparing the next generation of researchers and professionals to tackle these complex challenges. There is significant potential for this technology to be adopted by the commercial market, further protecting modern critical infrastructures against cyber-attacks, ensuring optimal control operation, human well-being and long-term national security.
Story by Krysta Murray
“Through our efforts, we hope to accelerate the adoption of renewable energy technologies and contribute to global efforts to combat climate change.” -Assistant Professor Adhikari