Linda Werner, an adjunct professor of computer science at UC Santa Cruz, has a knack for putting together teams of students that can tackle challenging engineering problems for their senior design projects. When researchers at IBM's Accelerated Discovery Lab in San Jose proposed a software project involving polymer chemistry, Werner assembled a multi-disciplinary team of seven students that has achieved impressive results.
The IBM Polymer Discovery team and other participants in the Baskin School of Engineering's Corporate Sponsored Senior Projects Program presented their work at Partner's Day on Thursday, May 29. This is the third year of the program, which gives engineering students the opportunity to work on real-world challenges proposed by corporate sponsors.
For the students on the IBM team, Partner's Day was just the latest opportunity to present their work. Two days earlier, they had presented a poster on their results at the Academy of Science and Engineering's International Conference on Big Data Science and Computing at Stanford University. And in April, they presented a paper at a conference of the American Society for Engineering Education (ASEE) in Connecticut, where they earned an honorable mention.
"It's been exciting," said team member Igor Shvartser, a math major with a minor in computer science. "Learning how to work with a team and with people from other disciplines is a key skill that we can take out of this experience."
The project also earned the team a Baskin School of Engineering Dean's Award. "This is the kind of experience that is possible if the students jump into the project and the company is really involved," Werner said. "When that's the case, we have a perfect situation."
The challenge for the Polymer Discovery team was to develop software for analyzing published information on the structure and properties of polymers. The ultimate goal is to enable the use of big data analytics to speed up the design and discovery of new polymers. But the diagrams of polymer structures in journal papers and patents, while easy for chemists to understand, are not in a format that computer software can read.
"Our task was to create a way to digitize those images so they can be stored in a database that researchers can query to find information about polymers quickly and easily," said team member Konstantin Litovskiy, a computer science major.
An open-source software tool called OSRA (Optical Structure Recognition Application) is able to do this for most chemical structure images. But OSRA doesn't work for polymers, which are represented using special notation to indicate substructures that are repeated multiple times in polymer chains. So the students developed P-OSRA, which extends the OSRA tool to support polymer diagrams.
"I'm very pleased with the progress they've made," said Julia Rice, a computational chemist and one of four IBM research scientists who worked closely with the student team. "The project was challenging because it required not only computer science but also an understanding of chemistry, and they had to figure out how to do it within the context of the existing software code."
When Werner was assembling the team in the fall, she was concerned about having computer science students work on such a chemistry-heavy project. So she went searching outside the Baskin School of Engineering to find a student with a strong background in chemistry. That's how Matthew Reed, a senior majoring in molecular, cell and developmental biology, came to be on the team.
"He was able to guide us through the chemical nomenclature," Litovskiy said. "Chemists and computer scientists communicate in different ways, so that was one of the challenges we had to overcome in order to collaborate as a team."
According to Werner, communicating with other disciplines is now a common challenge for software engineers. "Technology is everywhere now, so we're solving problems for other disciplines, and we need to be able to communicate with people in those disciplines," she said.
Werner said she was impressed by the ability of the students on the team to listen to and learn from each other. "They all had different interests and skills, and they were able to work together effectively," she said. "Aside from technical chops, the number one thing companies want is the ability to work collaboratively with others, and these students got a lot of experience with that."
Engineering dean Art Ramirez said the corporate sponsorships are especially valuable for students because they present real-world challenges. "The students who have participated in this program have been provided with a unique opportunity to experience working on real-world projects that involve design, budgets, deadlines, teamwork, and reviews with their team mentor," he said. "They have come away with a sense of professionalism and pride in their work, and we also take great pride in what the students have accomplished."
The Polymer Discovery team consisted of Thomas Goddard, Konstantin Litovskiy, Nathan Nichols-Roy, Matthew Reed, Igor Shvartser, Nicholas Smith, and David Zeppa. Their mentors at IBM's Almaden Research Center were Julia Rice, Hans Horn, Amanda Engler, and Jed Pitera.