Reporter with Silicon Hills News

Bob Metcalfe, professor of innovation at UT Austin talks with Andrea Thomaz, who teaches computer engineering and founded Diligent Droids, a healthcare robotics company.

Bob Metcalfe, professor of innovation at UT Austin talks with Andrea Thomaz, who teaches computer engineering and founded Diligent Droids, a healthcare robotics company. Photo by Hojun Choi

Entrepreneurs from around Austin gathered at WeWork downtown for the second stARTup Studio of 2016, and had a chance to hear about innovations being developed at The University of Texas at Austin.

The event, which began in 2013, is co-sponsored by WeWork, the Austin Chamber of Commerce and the Office of Technology Commercialization at UT Austin.

The meeting is meant to mimic an artist studio, where people can bring their work for others to critique, said Bob Metcalfe, professor of innovation at the Cockrell School of Engineering. He also serves as faculty director of the Innovation Center, which puts on the monthly stARTup Studio events.

On Tuesday evening, Andrea Thomaz, who teaches computer engineering, presented Diligent Droids, a company that seeks to provide aid to healthcare providers through robotic technologies.

“Being new to the Austin community, I was really looking to connect with potential strategic partners as well as people who can help advance the technology,” Thomaz said.

Thomaz said she hopes to apply her research in the field of artificial intelligence to help healthcare providers with their service.

“My work falls broadly into the world of artificial intelligence, but I’m interested in a really particular area,” Thomaz said. “I’m interested in social intelligence and how machines can fit appropriately into a human environment.”

Before joining the electrical and computer engineering department at UT Austin in January, Thomaz spent eight years researching and developing a robot that could learn from its surroundings.

“Social intelligence refers to the ability to react dynamically to the situation that humans are creating,” Thomaz said.

Recognizing the potential behind machines that can be taught to adapt to situations, Thomaz began researching how she could take the technology to market.

Thomaz said market research, which focused on healthcare providers, showed that nurses were often forced to spend a large portion of their time running around performing what Thomaz described as “hunting and gathering” tasks.

By providing hospitals with robots that can learn simple jobs, such as fetching warm towels or bringing supplies for procedures, Thomaz said nurses could spend more of their time tending to patients, and improving the quality of care.

Because every hospital functions differently, Thomaz said the machines would require time to “shadow” nurses to correctly learn their tasks properly.

“There is a lot going on in a hospital floor, and we need to have the robot adapt to how humans are currently getting a good job done,” Thomaz said.

The second presenter, Robert Crawford, professor of mechanical engineering, showcased his company, nVariate, with Ben Urick, a UT graduate student at the Cockrell School of Engineering. They said their service will help those who require computer-aided design, or CAD, for their businesses.

Crawford said computer-aided design currently faces many challenges in the realm of “surface-to-surface intersections” in which even the smallest gap or miscalculation can cause a huge design flaw.

For example, a design of a teapot may look good enough on a computer screen to have the design manufactured or created with a 3D printer. However, the teapot may have gaps that cause leaks due to the surfaces of the design not meeting at the seams correctly.

Urick, who also spoke to attendees at the event, said about 80 percent of the work that he performed as an engineer in the field required fixing these miscalculations, or “trim problems” by approximation and human calculation.

Rather than having to deal with human error and tedious calculations, Crawford and Urick said their process creates a “map” that can recognize flaws and fixes them in a way that does not require the manipulation of the design file itself.

“Our process identifies where these holes are, and it simply tells the graphic card simply not to show these parts,” Urick said. “Essentially, our technology identifies the math in the model, and does the rebuilding for you.”

Because each industry uses a different program to perform computer-aided design, Urick said nVariate’s process allows for designs to be fixed across these different platforms.

“A lot of times when you’re developing these technologies, you go down this ‘wormhole’ and you think everyone understands your industry, so it is really refreshing to hear from people who were not aware of these specific problems,” Urick said.