MIT’s disinfection robot: Using UVC radiation to eradicate COVID-19

    By Daniela Rus, Strategic Advisor


    As summer 2020 draws to a close, COVID-19 remains a major global health threat. In normal times, this is when more than 20 million students would be flocking to college campuses from coast to coast. But these are not normal times. The University of North Carolina-Chapel Hill and Notre Dame recently made headlines when, after welcoming students back for the school year, they were forced to revert to online-only classes when cases of COVID-19 climbed. With thousands of campuses scheduled to open in the coming weeks, including colleges, high schools, and elementary and middle schools, it is quickly becoming clear that social distancing, masks, and hand washing are insufficient to keep the virus at bay. What’s needed is a reliable way to systematically reduce the risk of exposure to COVID-19—at schools, offices, grocery stores, warehouses, and anywhere else the virus may be lingering and putting our health and our lives at risk.

    This is where technology can play a pivotal role, providing support to prevent the spread of the virus and the formation of hot spots using core-enabling technologies such as automatic, privacy-preserving contact tracing and robotic disinfection of public spaces. Here at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL), applying technology to help reduce the spread of COVID-19 has been the focus of some of our deployable research since the virus first emerged back in January. Today, our new UVC disinfection robot is ready to make the places where we live, work, and play much safer for us all.



    One of the biggest challenges in the battle against COVID-19 is that the virus is extremely effective at spreading. While medical researchers continue to learn more about the virus every day, what we know now is that larger droplets that are expelled when an infected person coughs or sneezes travel shorter distances and linger for shorter times, while the much smaller droplets that are expelled when an infected person breaths, speaks, or sings can remain stable for several hours in aerosol form. Importantly, all of these droplets can survive much longer (hours or even days) on certain surfaces. The CDC maintains a list of EPA-approved disinfectants that have been proven effective against COVID-19, but while it may be realistic to wipe doorknobs, faucets, and other high-touch surfaces within a household, larger spaces demand a faster, safer, more systematic approach.

    UVC radiation is one clear solution. UVC uses short-wavelength ultraviolet light to kill microorganisms and disrupt their DNA in a process called ultraviolet germicidal irradiation.

    Thanks to the spectral sensitivity of DNA, when the virus that causes COVID-19 is exposed to UVC light, the radiation can rupture the cell walls, making it impossible for the cell to replicate. While this approach doesn’t actually kill the virus, it essentially makes the virus powerless. But UVC light has a catch: it is dangerous to humans. For this reason, it had only been applied in controlled, isolated environments, such as to disinfect patient rooms, medical equipment, and surgical masks. Our goal was to expand the number of use cases by creating a way to safely apply UVC light over much wider areas. To do the job, we turned to a smart robot—one that is capable of operating without the presence of a human operator while keeping other human workers out of harm’s way.

    Ava Robotics emerged as our ideal partner. A spin-off company of iRobot, Ava Robotics specializes in combining robotics and AI to create mobile telepresence robots. Ava’s robots enable the operator to see and hear in a given space without the need to be physically present, making them a perfect fit for a UVC solution. Our team used Ava’s existing robot as a base, augmented it with a custom UVC module, and developed patrolling, coverage, and visualization algorithms for the robot to traverse the space and deliver a UVC dosage guaranteed to neutralize particles on surfaces and in aerosols. The UVC module replaces the telepresence video screen of the Ava robot with a high-powered UVC lamp that emits short-wavelength ultraviolet light to disinfect a given area. Because UVC disinfection is line of sight, algorithms are needed to drive the robot in such a way as to ensure that all the surfaces are disinfected. The robot is able to learn and create a map of a given space and then navigate from point to point. In this case, however, instead of cleaning house, our robot is delivering a dose of UVC light that is lethal to a deadly virus. To keep human workers safe from the UVC radiation, the dosage of light the robot emits is set to accommodate people as close as 15 feet. If the robot detects a person at an unsafe distance, it either moves out of range or tells the human worker to move, always giving the worker sufficient time to respond without harm.


    To put our creation into practice, we turned to the Greater Boston Food Bank (GBFB), an essential service that has been particularly hard-hit during the pandemic. The food bank’s warehouse was an ideal testing ground for the technology, and our robot passed with flying colors. During the initial tests, the robot was able to navigate among food pallets and expansive storage aisles across a 4,000 square foot warehouse. The robot was able to apply enough UVC light to neutralize 90% of coronaviruses on surfaces throughout this 4000 square foot facility in less than 90 minutes. By increasing the UVC dosage, we have the ability to increase the effectiveness of the disinfection process to 99% or even 99.9%.

    The GBFB quickly adopted the robot to disinfect areas such as the warehouse shipping floor—a particularly high-priority area where products are stacked to be picked up by 50+ trucks every morning for distribution across the Boston metro area. Even one contaminated area has the potential to spread the virus across the region. Because every product in the staging area is treated with UVC radiation before pickup, the food bank’s partners and distributors are able to operate with a high level of confidence that every shipment is virus-free.

    Our work, of course, is far from finished. Our team is now hard at work to refine and expand the robot’s capabilities, looking at ways to improve its onboard sensors to enable it to detect changes in its environment and adjust its speed to apply the correct dosage of UVC light to each new object or surface. We are also looking at methods that allow robots and people to be in the same space safely and harmoniously, and for ways to certify that all surfaces receive the right dosage of UVC light and are fully decontaminated. For larger spaces, the team is working to create teams of robots that use AI to communicate with each other to coordinate their efforts—all with as little human interaction as possible.

    Today, the list of potential commercial applications for CSAIL’s disinfection robot is growing quickly. The same UVC dosage used to neutralize COVID-19 can be used to neutralize a range of other pathogens. This opens the door to a host of additional applications to support industries such as food and hospitality, warehousing, and public transportation, as well as to clean offices, stores, and school campuses. In every public space, COVID-19 has made eradicating virus particles a priority. This new technology is quickly painting a new picture of the future—one in which we can all arrive at the office, grocery story, or classroom to find a space that has not only been dusted and swept, but is also certified to be virus free.


    About the Author

    A member of the ROBO Global Strategic Advisory Board, Daniela Rus is the Director of CSAIL at MIT, as well as a visiting fellow with The MITRE Corporation. Rus’s research interests are in robotics and artificial intelligence. She is the recipient of the 2017 Engelberger Robotics Award from the Robotics Industries Association, the 2018 IEEE Robotics and Automation Society Pioneer award, and the 2020 IJCAI John McCarthy award. She is also a Class of 2002 MacArthur Fellow, a fellow of ACM, AAAI and IEEE, and a member of the National Academy of Engineering and the American Academy of Arts and Sciences. Daniela earned her PhD in Computer Science from Cornell University.

    ROBO Global currently offers three innovative index portfolios, including our Robotics & Automation Index (ROBO)Healthcare Technology & Innovation Index (HTEC), and Artificial Intelligence Index (THNQ). To learn more, visit our website at


    Back to All Entries