Viterbi professors win Air Force’s Young Investigator Award


The mission of the U.S. Air Force is to fly, fight and win in air, space and cyberspace. Out of 280 proposals from researchers nationwide, two USC Viterbi School of Engineering professors received $450,000 grants from the Air Force Office of Scientific Research to aid the force’s mission through their investigative studies.

Qiming Wang (left) and Mahdi Soltanolkotabi (right), both assistant professors at the Viterbi School of Engineering, were presented with $450,000 in grants to further pursue their research. Photo courtesy of Valentina Suarez.

The grants, titled the Air Force Young Investigator Award, encourages creative research in science and engineering, which the professors will utilize to continue their studies and provide more opportunities to students.

“Winning awards is one thing, but I think the major goal for researchers is to publish high-impact papers,” assistant professor Qiming Wang said.

Through his research, Wang’s students are able to explore the mechanics of physical materials, specifically how soft surfaces like rubber, human skin and muscle can be deformed.

Assistant professor Mahdi Soltanolkotabi focuses on the intersection of data science, statistics and computer science through the creation of algorithms in machine learning.

“There are a lot of problems in engineering and basic sciences,” Soltanolkotabi said. “At the end of the day you want to optimize something. You want to maximize revenue, you want to minimize signal error.”

Wang’s research involves studying self-healing structures and how they are manufactured. He imagines that, through his work, fractured aircraft wings, bullet-damaged armor and broken cell phone circuits will be able repair themselves.

Wang creates smaller, durable yet lightweight models, such as cubes, capable of self-repair through 3-D printing. After these structures break, they mend themselves within hours almost to their original form, Wang said.

“We don’t need to spend extra money for [an aircraft wing],” Wang said. “If we have some material that can be made into [armor] and also can automatically, autonomously self-heal, then you can reuse it…”

These structures are made of a polymer-type material that Wang already developed in his prior research. When Wang first designed the material, he was able to infuse self-healable and photocurable properties into the polymer. However, mastering the coexistence of these two characteristics in the material’s design was a challenge.

“These two requirements compete with each other,” Wang said. “We have to find an adequate molecular structure to enable both behaviors.”

Although his current research is on a smaller-scale, Wang wishes to transition into large-scale printing and further expand on his studies in the future.

“We already have the material, [so] once we receive the grant … we can begin to do this 3-D printing of this kind of structure and test how this can help the repairing of the structures,” Wang said.

Soltanolkatobi’s research focuses primarily on learning data representations, which is the process of mapping or embedding data into representations that enable the computer to interpret high-dimensional data. 

These algorithms have many uses in modern signal processing and machine learning, applicable to artificial intelligence usage. For instance, they can solve word analogies, Soltanolkotabi said. 

According to Soltanolkotabi, optimization problems are commonly encountered in engineering and sciences, but for his current research, he decided to look at more unconventional problems. 

These problems have the potential for larger impact in signal processing and computational imaging. They will additionally decrease frequent purchases of expensive scientific equipment, as the computer will be able to augment many measurement devices.

“These algorithms can help overcome the deficiencies in many devices,” Soltanolkotabi said. “You can buy a cheap, small telescope but get the same image quality of a much larger more expensive one.”

Soltanolkotabi hopes that his research will reduce error and uncertainty surrounding modern artificial intelligence algorithms. Many of these algorithms fail because we do not have the technology required to further understand them, he said.

“These heuristics are like a black box and we don’t really understand why or how they work,” Soltanolkotabi said. “This lack of understanding could cause some issues and there are a variety of ethical concerns, fairness concerns … My hope is that by understanding the underlying theory behind why or how these algorithms work, I can make these inferential algorithms more reliable and fair.”