For Yung-Hsiang Lu, improving the energy efficiency of computer technology to provide real-world benefits has been a lifelong focus.
“When I was learning data structures, I began to see things from a different viewpoint—how to make things efficient,” says Lu, a professor of electrical and computer engineering and a university faculty scholar at Purdue University’s Elmore Family School of Electrical and Computer Engineering, in West Lafayette, Ind.
The IEEE Fellow’s research focuses on developing energy-efficient computing systems. Improved efficiency is increasingly essential for tasks like computer vision and imaging activities. It is particularly critical where battery weight, size, and capacity is a precious resource, such as in small and mobile devices like distributed sensor networks, autonomous robots, wireless communication, and real-time systems.
Over the past several years, his research has included collecting and analyzing data from networked cameras in department stores, optimizing how they place their products, and assessing COVID-19 lockdown compliance based on multinational camera data.
Lu is also guiding the next generation of researchers—both graduate and undergraduate students.
“I used to think I just wanted to write research papers,” Lu says. But after several of his students won the 2014 Schurz Innovation Challenge at Purdue, where students presented concepts for Web and mobile applications, Lu says he wanted to help students go from “research to technology transfer and commercialization. I want to see research results get used in the real world.
“I still do research, of course," Lu notes. "In fact, I received three research grants last year.”
Lu was inspired to conduct research in improving efficiency when he was an undergraduate student at the National Taiwan University, in Taipei, where his studies included courses in algorithms and data structures.
“Algorithms are very applicable to real life,” he explains. “I use them every morning to decide how I want to organize my day.”
After earning his bachelor’s degree in electrical engineering from the school in 1992, Lu went on to receive a master’s degree in EE in 1996 and a Ph.D. in EE in 2002, both from Stanford.
Lu has also written several books. His Intermediate C Programming (CRC Press, 2015) covers programming concepts, debugging, and the connection between programming and discrete mathematics. He also coedited Low-Power Computer Vision: Improve the Efficiency of Artificial Intelligence (Chapman and Hall/CRC, 2022), which collected methods for improving energy efficiency for computer vision on battery-powered systems.
Throughout the past decade Lu’s focus has been increasingly expanding beyond conducting research to include teaching real-world-relevant goals, helping students commercialize their technologies, and learning how to start and grow companies. He says he realized that he “wanted to help move our discoveries in the lab out to the world for impact through patents and technology transfer commercialization opportunities.
“For the 2014 Schurz competition, I’d given some suggestions to help the students develop the business plans and win the competition, and then helped them pursue commercialization, including developing SBIR [Small Business Innovation Research] proposals,” he says.
In 2015, Lu became a principal investigator for the U.S. National Science Foundation’s Innovation Corps (I-Corps) program, which helps NSF-supported professors understand the commercial value of their research.
That same year, Lu helped start the now-international IEEE Low-Power Computer Vision Challenge, an annual competition that aims to improve the energy efficiency of computer vision (CV) for running on systems with stringent resource constraints. Computer vision, Lu says, “remains one of the grand challenges in AI. To date, the competition has received more than 500 solutions for CV problems from over 100 teams around the world.”
Lu also helps undergraduates conduct research in relation to real-world problems through his involvement in Purdue’s Vertically Integrated Projects (VIP) program. For his work, Lu received the 2019 Outstanding VIP-Based Entrepreneur Award.
He has also been involved in several technology challenge events. He is one of the organizers of this year’s IEEE Autonomous Unmanned Aerial Vehicles (UAV) Competition, which challenges teams to see which of their UAVs can successfully follow a moving target without a teleoperator.
Lu was appointed the inaugural director of Purdue’s John Martinson Entrepreneurial Center in 2020—which, Lu points out, is one of many entrepreneurial programs and activities at Purdue.
One important observation Lu says he made during the NSF I-Corps program is that “the problems people in the real world face may not be the ones we at universities imagined.” He says it’s important for researchers to talk with people outside of academia because “they look at things differently.”
Lu also encourages students to try to get involved in a project that lasts more than one semester and for professors to provide these opportunities.
“When you are working on a one-semester project, it’s so short you don’t think about consequences for bad decisions,” he says. “If you are in a project for one year, a lot of bad decisions will come back and hurt you, which you learn from.”