UCSD duo takes aim at coronavirus with disinfection drones armed with UV lights
Two UC San Diego professors hope to turn a quickly growing hobby into an efficient, novel way to sanitize surfaces against viruses like the one that causes COVID-19.
Dr. Farshad Raissi, an assistant professor of cardiology, and Tara Javidi, a professor of electrical and computer engineering, independently came up with the idea to add ultraviolet lights to drones to clean items of the coronavirus, and then began working together when they realized, through a mutual contact, that they had the same goal.
“When COVID hit, I was thinking of using UV light to disinfect stuff around the house,” Raissi said. “Tara was training drones in her lab, using artificial intelligence to improve the drones’ flight.”
The two wondered about attaching UV light to a drone to quickly disinfect a room without exposing people to the virus or the UV rays.
UV-C, a shorter wavelength of ultraviolet light, is known to kill bacteria, fungi and viruses but also is known to damage cells, Raissi said.
Raissi said he and Javidi “talked about using a more practical source of the UV light, to use the drone as a means of delivering the disinfectant energy. It’s part of the project that the drone will learn to fly inside the space, whether a room or bus or metro car, and basically you don’t need to be inside that space to pilot it.”
UC San Diego says technology known as “nanosponges” developed by its engineers could work as a decoy to attract the coronavirus that causes COVID-19 and divert it from infecting human cells.
Flying a drone remotely has advantages beyond limiting exposure to danger, Raissi said. “The drone is hovering over [a] surface that may not get exposure with a stationary tower or bulky unit.”
In addition, Raissi said, a drone creates airflow, which may help move dust and other particles and allow the UV-C waves to better penetrate surfaces.
To ensure the drone isn’t weighed down by bulky UV light bulbs, Javidi attached LED strips with UV-C light. “They’re not as powerful,” she said. “Their efficiency is really low, so you have to think about optimizing the flight time and exposure.”
That’s something Javidi is working on in her lab. “It’s challenging,” she said. “Now you have to pilot the drone from the first-person view on the video.”
Optimizing the drone means improving the artificial intelligence. “Ideally,” Javidi said, “you want AI to be in a state … that you [are] able to put a drone in a space and let it fly and learn. In terms of AI, we’re not there yet. To build the AI engine, we need a lot of data.”
Gathering data, she said, is as simple as people who are “comfortable piloting their drones volunteering to help.”
“There’s not a lot of data about flying drones inside spaces, so that’s where peers can help develop AI,” Javidi said. “The idea is everybody should be able to do this with the drone they have” by attaching LED UV-C lights and flying it remotely to avoid exposure.
“We need people to share their flight training and experiences,” Raissi said.
The data also will help calculate “how much irradiation we need for a particular room. Everyone with the know-how of this platform should be able to use it,” Raissi said.
The next step for Raissi and Javidi is to determine how long a drone needs to fly for proper disinfection.
“We don’t know that answer, especially about this novel coronavirus,” Raissi said. “Compared to the older SARS virus or other RNA viruses, we have some yardsticks to compare. The exposure time to [kill] certain percentages of viruses is dependent on the amount of wattage per surface exposed. If the source of the light is close to that surface, more energy gets to that surface.”
Raissi said they plan to “send this to [Environmental Protection Agency]-designated reference labs for this type of testing, to see how much is needed for an average drone to hover around a surface to get rid of the virus.”
Javidi and Raissi are searching for funding to help that part of the project.
“You need a profile of effectiveness for us to be able to optimize the path of the drone,” Javidi said. “The studies we’ve seen on UV [efficacy] are too sparse. We need specific numbers on effectiveness.”
Noting that a drone with UV-C light has wide potential application, from hospitals and pharmacies to private homes and other spaces, Raissi said he’s hopeful about the possibility to safely rid surfaces of COVID-19.
“You can extrapolate,” he said. “If [studies show] you’re killing most of the viruses, the coronavirus should be mostly eradicated as well.”
For more information about disinfection drones, visit bit.ly/uvcdrone. ◆
Get the La Jolla Light weekly in your inbox
News, features and sports about La Jolla, every Thursday.
You may occasionally receive promotional content from the La Jolla Light.