CSL professor creates robotic avatar for hands-free medical care

3/25/2020 Lizzie Roehrs, CSL

Written by Lizzie Roehrs, CSL

In James Cameron’s blockbuster movie “Avatar,” humans explored a new planet without ever leaving their lounge chairs on Earth. CSL’s Kris Hauser and his team are creating a new type of avatar that could provide a similar ability to health care workers, who could treat patients remotely during a pandemic like the current COVID-19 crisis.

AVATRINA, the updated avatar version of an earlier project, TRINA (Tele-Robotic Intelligent Nursing Assistant), gives the user the means to interact with another human being in a “hands-free” fashion. AVATRINA allows the operator to move around, pick things up, and speak to other people without needing to be near the same room.

A group photo of CSL professor Kris Hauser and the students working on AVATRINA
A group photo of CSL professor Kris Hauser and the students working on AVATRINA

“In ‘Avatar’, the idea was that through some sort of neural link, a human would be able to control an alien’s body,” says Hauser, an associate professor of computer science. “We’re trying to do that through robotic technology.”

COVID-19 is believed to easily spread through person-to-person contact. While many working adults may elect to work from home, medical professionals have to be on site, which puts them at risk for exposure and transmission. Anyone needing treatment for other medical issues is also put at risk. Someone suffering from chest pains, for example, may be seen by the same emergency room staff as someone with advanced symptoms of COVID-19. The use of avatar robots would limit this face-to-face contact and could also slow the spread of future pandemics.

“The TRINA project got started in 2014 in response to the Ebola epidemic,” Hauser says. “We were interested in developing robotic solutions for handling infectious diseases by reducing the patient-to-caregiver transmission risk.”

Researchers have studied telerobotics for a long time, but there are numerous challenges in using technology as a replacement for humans, particularly in fluid and complex environments such as hospital rooms, which require dexterity and high levels of communication. AVATRINA incorporates artificial intelligence, virtual reality, and avatar technology to address some of these limitations.

Currently, the team is using virtual reality (VR) to allow an operator to feel as though they are in the remote environment where the robot is located. A user would put on a VR head set to view the environment and use two controllers to manipulate the movements of the robot and operate its arms and hands. The team is working on ways to transmit signals to allow the user to control the hands and movements of the robot in an intuitive fashion.

“AVATRINA combines a sort of Skype plus arms and wheels,” says Hauser, who said the applications could extend to telecommuting and e-tourism, among other areas. “The screen acts as your face and shows an image of the operator, and a microphone and speaker allows for bidirectional communication. There already exist ‘Skype on a stick’ telepresence robots that you can drive around. AVATRINA is the next wave in this technology that lets you touch and interact with other people beyond just speech.”

In order to make AVATRINA viable for the real world, developers face two significant challenges: a lack of skilled operators and the overall cost of the robot. According to Hauser, current medical tech like telesurgical robots require the operators to be trained for more than 10 hours before they can perform surgery on a living human being. To shorten this time spent in training, Hauser and his team hope to incorporate more artificial intelligence into AVATRINA to make interactions with the robot more natural and intuitive.

A photo depicting all of the labled parts of the AVATRINA robot
A photo depicting all of the labled parts of the AVATRINA robot

“We’re currently working with virtual reality, but operating the robot is a very foreign experience,” says Hauser. “There’s a fundamental disconnect between your body and senses and the robot’s arms and sensors. In the future, though, you may be able to operate these robots with a tablet or even a smartphone on just a point and click system.”

The second hurdle is cost. The robotic arms cost, at a minimum, between $10,000 and $20,000 each. Hauser says that they could use lower cost parts, but the overall quality and performance would suffer. The issue of cost, a total of about $100,000 per robot, is a problem for consumer use but is less of a problem when the tech is being used by hospitals that generally operate with a much larger budget than the average person. If put into mass production, Hauser says that cost per robot could drop.

Hauser and his team also plan to submit AVATRINA as an entry in the ANA Avatar XPRIZE competition, a competition to create robotic avatars held by the XPRIZE Foundation. Their team has passed the qualifying round, and over the next two years, this competition will hold its semifinal and final rounds. Hauser’s team is partnered with Duke University and VRotors, LLC to build and improve AVATRINA in the hopes of attaining the $5 million dollar prize.

For more information about AVATRINA, please visit http://avatarxprize.web.illinois.edu/ and http://motion.cs.illinois.edu/nursing.

             


Share this story

This story was published March 25, 2020.