Alumna Andrea Thomaz spends most of her work days watching a robot set the table or stack blocks.

Thomaz researches artificial intelligence and robotics. She directs the Socially Intelligent Machines Lab at Georgia Institute of Technology, which creates robots that communicate like human beings and learn from human demonstration. These types of robots have potential for future use in hospitals and businesses, where a person could train robots to carry out important tasks in fast-changing environments.

Although Thomaz’s lab is in Georgia, she spent her childhood in Texas. Born and raised in Houston, Thomaz earned a B.S. in computer and electronic engineering from UT. In 2006, she received her Ph.D. in robotics from MIT.

“I first got interested in studying robotics after taking a class in artificial intelligence at the very end of my time at UT,” Thomaz said. “I loved it because it was an area that combined knowledge about things like algorithms and speech recognition. The problems are all multifaceted.”

Thomaz specializes in creating socially intelligent robots. On Wednesday, she will give a talk at UT’s W. R. Woolrich Laboratories about her major project involving Simon, a learning robot. More Wall-E than the Terminator, Simon is a baby-faced, large-eyed robot that can learn simple sorting and household tasks from human teachers.

“Simon is meant to be an accessible service robot,” Thomaz said. “The idea is that, hopefully, in the near future, machines like Simon will be in businesses and homes, and people will be talking to [the robots] on the regular. We observe how successfully [the robot] learns from people and then adjust the algorithms.”

Part of Thomaz’s challenge is achieving more human-like communication with robots. When creating dialogues between people and robots, she said she has to consider the human side of the equation along with the robot’s.

“You need to have a good understanding of human speech patterns and teaching methods,” Thomaz said. “You need to consider things like the natural movements that people make when they speak and how that affects communication. All of that needs to come together.”

Sonia Chernova, director of the Robot Autonomy and Interactive Learning Lab at Worcester Polytechnic Institute, worked on researching robot learning techniques with Thomaz. Chernova said Thomaz’s research makes robots approachable to the average person.

“[Thomaz] is one of the pioneers in research on robots learning from demonstration,” Chernova said. “Her research is exciting because it creates the opportunity to empower everyday people to interact with some of the most advanced technologies that we have today.” 

As a leading researcher in a traditionally male-dominated field, Thomaz has noticed a gender gap in the tech world. But Thomaz thinks that the issue can be addressed through more outreach to women interested in tech.

“There are definitely more men than women working in robotics right now,” Thomaz said. “But diversity shouldn’t always have to be a problem.”

Thomaz participates in several ventures to support women in tech. She mentors female Ph.D. students and teaches workshops targeted at
female researchers.

“I think it’s just a matter of time before we see critical numbers of women studying robotics,” Thomaz. “There are already so many great women doing wonderful things in research and becoming very successful.”

Thomaz said the path in the tech world doesn’t always have to be straightforward. She said her diverse educational background contributed to her love for the multidisciplinary area
of robotics.

“I think that it’s okay to try new things and change [your] mind about where [you] want to go in life,” Thomaz said. “I was lucky to discover my passion for robotics.”