By Tom Dinki
Published July 28, 2023
Five faculty members in the School of Engineering and Applied Sciences (SEAS) have earned National Science Foundation CAREER Awards for research that could protect data in household appliances, catch cryptocurrency money launderers, extend the life of cars and engines, improve artificial intelligence-assisted health care and advance our understanding of life itself.
Mingchen Gao, A. Erdem Sariyuce and Ziming Zhao, assistant professors in the Department of Computer Science and Engineering; Prathima Nalam, assistant professor in the Department Materials Design and Innovation; and Sangwoo Shin, assistant professor in the Department of Mechanical and Aerospace Engineering, are recipients of one of the most prestigious honors for early-career scientists and engineers.
In total, they will receive nearly $2.9 million in funding for their respective projects, which include both research on pressing societal problems and outreach to diverse communities.
“I am extremely proud of the research that our early career faculty are leading,” says Kemper Lewis, dean and professor of the School of Engineering and Applied Sciences. “As a School of Engineering and Applied Sciences, we take immense pride in the profound impact our faculty have on society, and we remain committed to fostering an environment that cultivates innovation, collaboration and societal responsibility.”
Advancing AI-powered medical diagnosing
What if a patient could get an accurate assessment of a rash on their skin by simply pulling out their phone and taking a photo of the infected area?
This is the future that Gao aims to create through the advancement of AI-powered medical imaging diagnostics. Funding by a $579,000 grant, she will create algorithms that help machine learning models analyze medical images, leading to improved clinical decisions and more confidence in AI-assisted health care.
“Life-threatening diseases could be caught early or avoided entirely,” says Gao.
Existing AI-enabled medical devices use deep learning models to scan complex patterns in large data sets and flag suspicious findings on a patient’s scan as capably as board-certified radiologists. However, says Gao, deep learning is heavily dependent on large, high-quality labeled datasets only available in lab settings.
Gao aims to improve the performance of deep learning models in clinical environments, where data may be limited, patient populations may be more diverse, and rare diseases cannot be ignored, as in simplified lab settings. Most crucially, she will train the models not to forget previous information upon learning new information, a problem known as catastrophic forgetting.
The research will be conducted using chest X-rays and eye images supplied by Cornell University and the University of Virginia. The project will also involve a diverse cohort of student researchers, as well as work to help ensure consistent screening services for glaucoma are provided in underserved communities.
Catching cybercriminals and money launderers
Networks that transfer money and messages are all around us. And they are open to attacks.
With a $556,000 grant, Sariyuce aims to create a novel approach to network analysis that could better detect cyberattacks and connect the dots between money launderers.
“Money transfers, communications between computers and face-to-face interactions are all networks,” says Sariyuce. “So, analyzing networks is critically important for bank fraud detection, ensuring cyber-secure environments and even preventing pandemics.”
Sariyuce’s research will focus on temporal networks, which contain links that are only active at certain points in time. Their temporality amplifies the diverse nature of networks, obscures the notion of scale and explodes the size, he says.
The research could aid the detection of money laundering through cryptocurrency.
“Cartels love Bitcoin,” Sariyuce says. “Cryptocurrencies create an illusion of privacy where we don’t know who is who but see all the transactions. Temporal motif-based methods can offer a systematic way to detect anomalies in which coordinated illicit activities occur.”
The project will also help train the next generation of computer scientists from diverse backgrounds. Sariyuce will organize computer science workshops for high school students from Hispanic, Burmese and Somali communities in Buffalo with the help of community organizations.
Protecting smart appliances from hackers
There are an estimated 17 billion smart devices worldwide, ranging from self-driving cars to smart refrigerators and thermostats. However, these devices are often vulnerable due to flawed security and attackers are increasingly leveraging these weaknesses to infiltrate corporate networks, according to Microsoft’s Digital Defense Report 2022.
Zhao received a $565,000 award to discover and correct the vulnerabilities in trusted execution environments (TEEs), a core component of Internet of Things (IoT) device security.
“The vast number of these devices and their crucial role in our daily lives make it imperative to prioritize their security,” says Zhao, who joined UB in 2020.
TEEs are protected modes on the main processor that execute sensitive code and store sensitive data like credit card and medical information in isolation, so that an attacker can’t access a device’s data even if they gain access to the main operating system.
In addition to improving TEEs security, Zhao also aims to advance the education pedagogy of Internet of Things security.
“The project's broader significance and importance, beyond securing the IoT infrastructure, are to train the next generation of cybersecurity researchers, educators and practitioners with deep theoretical understandings and practical skills in this field,” he says.
Lubricant for a longer-lasting car
Wear caused by friction is a widespread problem in machines, leading to shorter lifecycles, increasing waste and a growing environmental burden.
“In the automotive industry alone, nearly 5% of global energy consumption is a result of energy loss due to friction,” says Nalam.
To address these challenges, Nalam will use a $667,000 grant to develop efficient, low-pollutant lubricants. She will investigate the physical behavior and viscoelastic properties of two-dimensional materials, such as atomically thin layers of graphite and molybdenum sulfide, as potential lubricant additives to protect engine walls. The layered structure of the materials would allow the intercalation, or insertion, of small molecules such as oil between its layers to better lubricate surfaces.
The results will advance the application of 2D materials as coatings to reduce friction loss—and, therefore, wear and tear—in gears and engines.
“Exploring liquid behavior within nanoconfined spaces holds the promise of unlocking a novel ability to manipulate material surfaces,” says Nalam. “This presents significant opportunities in energy storage, catalysis and the development of super lubricious surfaces.”
The project will also provide summer research training for underrepresented students, fostering skills that enable the acceleration of material discovery and development. The research will promote dialogue and outreach in the tribological (science of wear, friction and lubrication) community as well.
Decoding life though cell movement
Cellular organisms and cell parts relentlessly move around the body to perform various functions. Understanding this movement could provide insight into how wounds heal, pathogens are transmitted and cancer spreads. It can even lead to new approaches for treating wastewater and helping therapeutic drugs reach their targets.
Shin was awarded $500,000 to explore one important aspect of cells that affect their movement: lipid bilayers, or the thin membranes encapsulating the cell that control the entry or exit of water and ions.
“By delving into the fundamental physics governing the motion of these microscopic entities, our research has the potential to advance our understanding of life itself,” says Shin.
The study will synthesize lipid vesicles—cell-mimicking droplets that are encapsulated by lipid membranes—and observe their motion in various conditions under a microscope.
The project will also promote diversity and inclusion in the science, technology, engineering and math fields by providing research opportunities to undergraduate students from underrepresented groups. The aim is to motivate more underrepresented students to study soft matter physics and pursue graduate degrees.
Visit the SEAS website to learn more about CAREER Award recipients.