Academic Year Highlights 2022-2023

A world of ethical autonomous systems

Houssam Abbas, assistant professor of electrical and computer engineering, received a Faculty Early Career Development, or CAREER, award from the National Science Foundation, including a grant of almost $500,000 over five years. Abbas aims to advance the engineering and scientific discipline of computational ethics, used in designing autonomous systems — such as self- driving cars, unpiloted aerial vehicles, and assistive robots in medical facilities — that interact with people daily. He also understands that what is considered ethical may vary among different communities.

“It is not enough that these systems are probably safe. We also expect them to abide by the community’s ethical principles,” Abbas explained.

For instance, a self-driving car may encounter a situation where it must make an ethically fraught decision: with no alternatives, does it run into a fence and potentially injure its passengers, or run into a pedestrian? A process to model, verify, and analyze autonomous systems’ behaviors in such situations is crucial. Abbas will create and develop engineering tools allowing autonomous system designers to formalize, program, and verify the application of ethical principles.

“When philosophers and policymakers debate these questions, there will be a way to experiment with the consequences of different ethics rule sets,” Abbas said.

Shaping the future of online engineering education

In 2021, the College of Engineering collaborated with Oregon State Ecampus to launch the Center for Research in Engineering Education Online, or CREEdO, to support research that informs the development of online engineering programs.

The center funded three projects in its inaugural year. One, led by Margaret Burnett, distinguished professor of computer science, leverages the GenderMag system she developed to identify and fix gender-inclusiveness “bugs” in software. Burnett and her co-investigators, Anita Sarma, associate professor of computer science, and Lara Letaw, computer science instructor, will help Ecampus faculty create inclusive engineering courses that integrate GenderMag concepts into their online curricula. The project will create a path for the College of Engineering to become a leader in inclusive online education.

Another project, led by Raffaele de Amicis, associate professor of computer science, will eval- uate the integration of extended-reality simulations and learning tools into engineering courses. XR simulations have been suggested as a way to enhance conceptual thinking in engineering. However, there’s scant rigorous identification of the requirements necessary to implement them successfully across engineering disciplines.

De Amicis and his co-investigators, Yelda Turkan, associate professor of geomatics, and Onan Demirel, assistant professor of mechanical engineering, hope to bridge that knowledge gap.

The results could represent a significant contribution toward understanding how XR solutions can be used in engineering education.

In the third project, Jeff Nason, professor of environmental engineering, will assess the use of virtual laboratories for conveying engineering practices that are not easily accomplished through in-person university labs. His project team will partner with Oregon State instructors to deliver both the physical and virtual versions of a laboratory, comparing student motivation, epistemic practices, and development of an engineering identity.

The expected outcomes will include the development of partnerships outside of Oregon State and a proposal for external funding to further develop and expand the research. Nason’s project, for example, has already earned a two-year $200,000 National Science Foundation grant.

Creating energy storage solutions

Yue Cao, assistant professor of electrical and computer engineering, has received a Faculty Early Career Development, or CAREER, award from the National Science Foundation, which includes a grant of almost $500,000 over five years. Cao’s research utilizes what he calls “virtual” systems — such as water heaters and HVAC — often over-looked as alternatives to traditional energy storage.

“Storing energy is not their primary purpose, but they are tied to the grid or other energy resources,” Cao said.

Cao’s research will help create a universal equivalent circuit for multiple energy storage systems, controlled by connected power electronics. He will develop a design approach to optimally size hybrid energy storage systems, increasing their life and dependability. His idea is to modulate grid energy usage by dynamically regulating virtual energy mass.

“If I have rooftop solar panels, and a cloud blocks the sun, solar power will be reduced,” Cao explained. “Current systems use grid power to keep the air conditioner running. With an integrated energy system, however, power used by the air conditioner — the virtual resource — could be adjusted to match the reduced solar power.”

Currently, Cao’s projects involve energy storage problems, including fast charging stations for heavy-duty trucks on rural highways, electrification of locomotives, and wave energy.

Cao and Alan Fern, professor of computer science, are working with Daimler Trucks of North America to develop extremely efficient and reliable power electronics and motor drive technology as well as innovative artificial intelligence-powered energy management tools for a fuel-cell heavy-duty electric truck dubbed the ‘SuperTruck 3.’

The U.S. Department of Energy awarded Daimler $26 million in November to develop the Class 8 fuel-cell truck with 600-mile range and 25,000-hour durability. Cao and Fern have been awarded$860K to work on the project.

Cao and Alan Fern, professor of computer science, are working with Daimler Trucks of North America to develop extremely efficient and reliable power electronics and motor drive technology as well as innovative artificial intelligence-powered energy management tools for a fuel-cell heavy-duty electric truck dubbed the ‘SuperTruck 3.’ The U.S. Department of Energy awarded Daimler $26 million in November to develop the Class 8 fuel-cell truck with 600-mile range and 25,000-hour durability. Cao and Fern have been awarded$860K to work on the project.

Oregon State is part of a $20 million federal effort to develop artificial intelligence for agriculture. Launched by the National Science Foundation in 2021, the AgAID Institute will tackle mounting challenges — such as diminishing water and labor supplies, weather varia- tions, and climate change.

“It is essential to improve the robustness, efficiency, and adaptability of food production,” said Alan Fern, professor of computer science. Fern serves as the institute’s principal investigator for Oregon State, heading up an AgAID team of 13 researchers in the College of Engineering. “The institute aims to achieve this by identifying the best ways to integrate humans and AI/robotics technology.”

Oregon State’s team will lead the institute’s fundamental and applied research in AI, robotics, and human factors. The group will work closely with agricultural researchers and others, including farmers, ranchers, and agricultural commodity producers. Researchers will explore robotics and intelligent systems holistically and examine their impacts on people.

“Humans and AI/robotics have very different capabilities and competencies, and there are many possible ways of combining them for any given agricultural application. But only some of those combinations will be effective,” Fern said. “We want to design and build workflows that have real utility and thus actually get used.”

Learning without supervision

Xiao Fu, assistant professor of electrical and computer engineering and artificial intelligence, has received a Faculty Early Career Development, or CAREER, award from the National Science Foundation. Fu will use his five-year, $500,000 award to develop a suite of nonlinear factor analysis tools and contribute to a deeper understanding of unsupervised machine learning and sensing systems.

Factor analysis tools are cornerstones of many sensing and learning applications, such as document analytics, brain signal processing, and representation learning. They’re designed to detect meaningful information hidden in large data sets.

One of Fu’s goals is using nonlinear factor analysis to understand and advance unsupervised deep representation learning, which is considered a critical tool to alleviate the high demand for labeled data in modern AI systems.

Supervised machine learning algorithms learn through exposure to labeled inputs that correspond with specific outputs. But the training process can be costly and time intensive, because reliable data annotation must be done by experienced workers.

Fu hopes to “reverse engineer” the data generating/acquisition process, so that machine learning and sensing algorithms can recognize and categorize unlabeled data — images, for instance — without being trained, by identifying and interpreting essential factors hidden within the data.

“I’m envisioning important theoretical advances and breakthroughs to understand this inverse learning process that does not require labeling,” Fu said. “From there, I might be able to build systems with many new functionalities.”

Teaching encryption as a safeguard of civil liberties

Glencora Borradaile, professor of computer science, has published “Defend Dissent,” an introductory textbook on digital security emphasizing the role of encryption technology in protecting civil liberties. Although Borradaile uses the text in a course they teach, the material, drawn from examples of U.S. social movements, is intended to be accessible to any curious person.

“There wasn’t a good single resource written at the right level,” Borradaile said. “There are training materials by the Electronic Frontier Foundation and the like that aren’t deep enough for a university course, and there are cryptography and security texts that aren’t appropriate for my nonmajor, freshman-level audience. I also wanted to make sure it would be of interest to nonstudents.”

Borradaile’s Communications Security and Social Movements course has proven popular with students in both technical and nontechnical majors. The course motivates learning about communications security by including historical and contemporary examples of surveillance used to hinder social movements, such as the civil rights movement.

“Defend Dissent” was published as an open, English-language text-book and is about to be released in Spanish. A digital copy may be downloaded free from open.oregonstate.education/defenddissent. 

Garbled circuits for secure computing

Upon completing his homeschool education in Corvallis at 18, Lance Roy took an unusual step — directly into the College of Engineering’s doctoral program in computer science.

Roy, now 23, presented research findings at the virtual International Cryptology Conference last summer with Mike Rosulek, associate professor of computer science. Roy and Rosulek have developed a secure computation protocol implementing garbled circuits, achieving 25% higher efficiency than Rosulek previously thought possible. Their protocol saves time and reduces energy costs for groups needing to compute together while maintaining individual data privacy. Rosulek deemed Roy the mastermind behind the garbled circuits.

“Mike was excited when I showed it to him. During winter 2021, we refined the technique and wrote up the result,” Roy said.

Roy’s work is funded by a Department of Energy Computational Science Graduate Fellowship, a program enabling him to focus on fundamental science and engineering applications.