UA Licenses Deception-Detecting AVATAR to Startup

Jay Nunamaker developed the AVATAR system with graduate students to make border crossings both more secure and more efficient.

Jay Nunamaker developed the AVATAR system with graduate students to make border crossings both more secure and more efficient.

(Photo: Paul Tumarkin/Tech Launch Arizona)

The University of Arizona has licensed a deception-detection technology to local startup Discern Science International Inc. The development of the technology, known as the Automated Virtual Agent for Truth Assessments in Real-Time, or AVATAR, was led by Jay Nunamaker, director of the UA Center for the Management of Information, who also serves as president and CEO of the company.

The AVATAR system works like this: As users answer interview questions posed to them by an interactive electronic interviewer, the system records facial expressions in high-definition video. At the same time, its many sensors measure and record thousands of signals from the subject's voice, body and eyes. All of this information is routed through a complex analytical algorithm, and the results are produced almost instantly: Green means the subject is clear to pass, yellow means there are some issues to be investigated, and red means there are serious issues that require deeper investigation.

Nunamaker, UA Regents' Professor and Soldwedel Chair in Management Information Systems, originally conducted the research that led to AVATAR's creation at the UA Eller College of Management's MIS program through the Borders Research Project. The Department of Defense, National Science Foundation, FRONTEX and Department of Homeland Security funded the work.

"The AVATAR technology was developed for the most rigorous and difficult task of detecting deception at the border," Nunamaker said. "It had to allow innocent people to quickly pass through the border, while simultaneously identifying those attempting deception — and do it all in about 30 seconds."

Past research has shown that the system is 70 to 92 percent accurate, depending on the application. In comparison, many studies show that humans can catch deception only 54 percent of the time. Like other types of artificial intelligence, the AVATAR can learn and will improve over time.

"We have designed studies to accelerate the AVATAR's learning by engaging people who will be well compensated if they can fool the AVATAR," Nunamaker said.

While the initial applications have been around border security, the company's plan is to commercialize the AVATAR technology across a broad range of fields. Tech Launch Arizona, the UA office that commercializes inventions stemming from university research, worked to protect the intellectual property and then collaborated with Nunamaker and David Mackstaller to license it to the new company.

"I continue to be excited by the broad range of intellectual property related to the fourth industrial revolution being developed at UA," said Doug Hockstad, assistant vice president for TLA. "The research being performed at MIS, and across campus, is presenting more and more opportunities for UA to have an impact on the world. We're excited to be partnering with Jay and the Discern Sciences team to contribute to a more secure society."

Many people collaborated with Nunamaker over many years to create the AVATAR — most notably Judee K. Burgoon, professor and director of human communication research for the UA Center for the Management of Information; Douglas C. Derrick, a former graduate student in Nunamaker's lab who is now an assistant professor of IT innovation at the University of Nebraska at Omaha; and Aaron C. Elkins, a former graduate student who is now director of the artificial intelligence program at San Diego State University.

"We are already talking to TLA about licensing additional technology to package with the AVATAR," said Nunamaker. "This way, we can continue to focus on our research knowing that we have processes and people who stand ready to help get our ideas out into the world where they can make our world a little safer."

A version of this article orginally appeared on Tech Launch Arizona's website: