Harry Gao '11 to receive Thomas Jefferson Prize in Natural Philosophy
The 2011 winner of the Thomas Jefferson Prize in Natural Philosophy never liked to read and admits that even now he doesn’t read anything that he doesn’t absolutely have to.
But there was one series of books that Harry Gao ‘11 lived vicariously through, one character whose existence harmonized with Gao’s overwhelming passion to understand how and why things do what they do, and how they can be made better.
Sherlock Holmes.
“I really love those detective things and that mystery solving,” he said recently. “It was the only book I could not put down.”
Gao, who carries a double major in computer science and mathematics, is the first computer science major to earn the Jefferson Prize, which will be given during Charter Day ceremonies on Feb. 4 in recognition of excellence in the sciences. The prize honors the relationship that Thomas Jefferson enjoyed with Professor William Small, his College tutor in mathematics and the natural sciences.
“He’s fearless,” Associate Professor of Mathematics Michael Lewis said of Gao. “Some students see how hard a problem is and back off. Harry is not put off by how hard problems are. He goes right at them, and that makes him great to work with.”
The senior was recently selected as a finalist in the Computing Research Association’s Outstanding Undergraduate Researcher Award competition for 2011, which speaks to his capabilities as a researcher.
Only 10 male undergraduate students were selected as winners, runners-up, and finalists from a pool that included all universities in North America.
“Harry is the most motivated student among all of the undergraduates who have worked with me,” said Qun Li, graduate director and associate professor of computer science. Li was Gao’s freshman advisor, and laid out a plan for him that involved doing as much research as possible, as soon as possible.
“He is definitely one of the best students I have met in my six years on the faculty of the College,” Li continued. “He always tries his best in seeking all opportunities and making things happen.”
With a chuckle, Gao admits his unbridled curiosity created its share of family drama early on. He almost caused his grandparents’ home in China to burn down when he was a kid. And he remembers his mother’s face when she handed him his first computer and owner's manual then watched as he blithely typed in one “fuzzy computer word” after another in an English language he did not yet understand then hit ‘command’ just to see what would happen.
“Yet they were actually very encouraging to me to experiment with things,” he said, slightly amazed. “I was always wondering ‘Why does this work?’ and ‘How exactly does this thing do what it does?’ ”
Gao has been a research warrior since coming to W&M from Falls Church, where his family ultimately settled once leaving Jinan, China. His current project is in conjunction with Ph.D. student Wei Wei on the “distributed consensus” problem.
The problem is to make multiple servers (such as gmail servers) distributed in a large area network (over several continents) both consistent in data replication and efficient in user response.
“Over time I discovered that Harry is quite good both at algorithm design and on research problems with a more theoretical flavor,” Li said. “That is why I started to challenge him with more advanced topics and harder problems. In the past several years, he has been exposed to a few research problems ranging from protocol design to system implementation to theoretical analysis. The results he achieved are deep and broad.”
The findings of his earliest project, developing a more reliable and secure protocol for gathering information from roadside sensors, were published in the International Conference on Wireless Algorithms, Systems, and Applications and the Journal of Communications.
His second project was funded in summer 2009 by a National Science Foundation Computational Science Training for Undergraduates in the Mathematical Sciences grant. It was aimed at improving the function of sensors that detect human motion. These sensors are often used in senior care to alert care providers if a person moves or falls. To reduce battery drain, the sensors are equipped with a sleep mode that kicks in when the sensor remains inactive for a period of time.
“The problem is the sensor can’t detect motion when the sleep mode is on,” Gao said.
Working with fellow student Andrew Wilcox, they reduced battery drain and eliminated the need for a sleep mode by redesigning the data sent by the sensor to use less information. By compressing the data into one small byte, less of the battery was used.
Most recently, Gao has examined the secure use of inexpensive RFID technology (the same system used for collecting tolls with a “smart tag” EZ Pass) to provide, among other things, proof of location for people.
Perfected, it would help suspects in criminal investigations establish an alibi as to their whereabouts. But it could also be used for the more mundane, such as allowing people to board the subway without having to have the proper change or slip a card into a slot.
“These sensors are really cheap, between a penny and three pennies apiece, and they are expected to become even cheaper,” Gao explained. “You could put thousands of them on the ceiling or at the entrance -- at a cost of $10 or $20. As you walked in, your cell phone would automatically register with a tag, and your credit card could be billed once a month for your usage.”
At this stage, Gao calls his ideas and the research behind them “scratching on the back of an envelope.” Whether it’s feasible from an engineering standpoint remains to be seen.
“When I start to research something, I try to have usages in mind,” he said. “I don’t want to get too academic, too far away from the real world.”
The idea of building a career making life a little easier for people to live has crossed his mind, but that’s down the road. Marriage and graduate school loom on the immediate horizon.
“I’m quite happy putting (an idea) out there,” he said. “I don’t really want to think about the direct financial incentive to it. If it’s actually used there will probably be something in it for me. But that’s not my major concern right now.”
