UV disinfection is one of the few areas where autonomous robots can be immediately and uniquely helpful during the COVID pandemic. Unfortunately, there aren’t enough of these robots to fulfill demand right now, and although companies are working hard to build them, it takes a substantial amount of time to develop the hardware, software, operational knowledge, and integration experience required to make a robotic disinfection system work in a hospital.
Conor McGinn, an assistant professor of mechanical engineering at Trinity College in Dublin and coleader of the Robotics and Innovation Lab (RAIL), has pulled together a small team of hardware and software engineers who’ve managed to get a UV disinfection robot into hospital testing within a matter of just a few weeks. They made it happen in such a short amount of time by building on previous research, collaborating with hospitals directly, and leveraging a development platform: the TurtleBot 2.
Over the last few years, RAIL has been researching mobile social robots for elder-care applications, and during their pilot testing, they came to understand how big of a problem infection can be in environments like nursing homes. This was well before COVID-19, but it was (and still is) one of the leading causes of hospitalization for nursing home residents. Most places just wipe down surfaces with disinfectant sometimes, but these facilities have many surfaces (like fabrics) that aren’t as easy to clean, and with people coming in and out all the time, anyone with a compromised immune system is always at risk.
Akara Robotics says radiology rooms are an ideal place for its robot to work right now. The rooms tend to be small and uncluttered, making robot navigation easy. In addition, the radiology machines are expensive and are often cleaned by radiologists, which not only takes up time that a specialist could be using to help patients but also puts the specialists themselves at risk.Photo: Akara Robotics
“UV seemed to offer a lot of potential for addressing this problem,” McGinn told us last week. “It’s something that covers a lot of space, and it’s something that can very easily be put on a robot.”
The researchers thought about developing this concept further, but after a bit of poking around, it turned out that while there was plenty of data showing that UV light does kill viruses and bacteria, there wasn’t a lot of useful design information, like what kind of light you need, how powerful it has to be, and how long you have to illuminate difference surfaces from specific distances. McGinn says that as a sort of a side project, his group started working with microbiologists at Trinity to figure out all of these parameters. “The idea was that if we came up with an effective way to make it work, we could always add it to our robot.”
Meanwhile, McGinn’s team kept working on the social applications for their robotic platform, and last year, formed a company called Akara Robotics to launch Stevie, a mobile social robot for elder-care facilities. “There was a lot of interest in us, and we’re making progress, and then this COVID thing came about and so we had to reprioritize.” McGinn contacted some folks at the Irish national health service, the HSE, to discuss how robots might be helpful, and from a bunch of ideas that included telepresence and delivery, a UV disinfecting robot was what really got their interest. “This conversation took place on Wednesday [11 March] at 6 or 7 p.m., and they said, ‘Can you build a proof of concept and bring it back here for a demo at lunchtime on Friday?’ ”
The Akara team took all of the UV disinfecting research they’d been doing low-key work on and combined it with the navigation stack (and other autonomy software) that they’d developed for Stevie. Then they crammed it all into a super-low-cost and development-friendly platform they happened to have lying around that they could quickly iterate.
“The fastest way to test the principle was to make adaptations to the TurtleBot 2,” McGinn says. “It’s a delightful platform to work with because there’s so much software in the box, and it’s open source and easily modified, and was designed around the fact that people will need to prototype things in hardware.” In about 24 hours, McGinn and his colleagues were able to take the TurtleBot that was being used by one of his Ph.D. students for an entirely different project, mounted the UV light and computer on it with an inverter and some extra batteries, and they had something they could bring to the hospital to test.
The robot was functional by Thursday night. Akara spent Friday morning tidying up the robot to make it look presentable, and they made it to the demo on time.
Despite the speed with which it was constructed, the prototype has held up well through rigorous in-hospital testing. “We’ve now brought it to two different hospitals, and one of them we’ve been going to every week for the last three weeks,” McGinn says. “We’ve also tested it in the lab a bunch, and even brought it on public transport, and we’ve had no failures that we haven’t been able to fix in just a few minutes. It might look a bit wobbly in the videos, but it’s surprisingly robust.”
Rather than try and change existing decontamination workflows in most of the hospital, Akara is instead focusing on areas where different infection-control procedures mean that there’s an opportunity to jump in and make things safer and more efficient without requiring a lot of disruption. “What we’ve been trying to do with the relatively limited resources we have is to find out what is the most amount of impact we can deliver in the fastest amount of time with the least amount of effort,” says McGinn. “That’s been what our critical path has been for the last couple weeks.”
It turns out that radiology is the best place for the robot to work right now, for a variety of reasons. First, radiology rooms tend to be small, uncluttered, and windowless, which makes robot navigation easy. Second, since the imaging machines are expensive, they’re often cleaned by radiologists, which not only takes up time that specialists could be using to help patients, but also puts the specialists themselves at risk. And finally, because the machines need cleaning after each COVID patient, traditional cleaning methods reduce the daily capacity of the machine by a factor of five or six. If you can drop a cleaning robot in there that speeds things up so that you can sneak in even one more patient per day for a CT scan, that robot will quickly pay for itself.
This video is from late last week, and shows the robot (named Violet) autonomously cleaning a radiology room at the Midlands Regional Hospital in Tullamore. The total cleaning time is less than 15 minutes, and the only required human involvement is a wipe-down of the robot itself.
Unlike other UV disinfecting robots we’ve seen, Violet uses one single UV light that it has to point directly at things it wants to disinfect. Fortunately, coronaviruses are relatively vulnerable to this method, and the light is both effective and safer for humans to be around. Part of the goal of the testing that Akara is doing right now is to confirm that Violet’s UV system does in fact disinfect like it’s supposed to (and testing so far has shown that it works), as well as establishing safe operating procedures for its users.
Conor McGinn and Niamh Donnelly, co-founders of Akara Robotics, during a recent test of their UV disinfecting robot Violet at a hospital in Dublin.Photo: Conor McGinn via Twitter
With a working prototype, the question now is what happens next. As far as the hardware is concerned, McGinn says that the TurtleBot platform was never intended to be more than a test-bed and that Akara has been putting together a design for a “substantially better robot.” However, building a new robot takes funding, and getting VCs to back hardware on short notice is a challenge, especially right now. Akara is applying for a number of grants in the meantime, but with hospitals needing solutions immediately, McGinn says that “where time is of the essence, TurtleBot delivers a huge amount of value.”
McGinn tells us that Akara has local clients that are very interested in Violet, both for coronavirus in the short and medium term, and general disinfection in the longer term, since UV disinfection will still have value pretty much anywhere that people are particularly at risk from infectious diseases. The idea is to offer the robot as a service, and while McGinn says it’s a bit early to talk about costs, he expects that Akara’s robot will be significantly cheaper than other robots with similar capabilities currently on the market. Meantime, they’re continuing to test Violet in hospitals as much as they can, in the hope that funding comes through to help them start scaling up and helping people as soon as possible.
[ Akara Robotics ]
Evan Ackerman is a senior editor at IEEE Spectrum. Since 2007, he has written over 6,000 articles on robotics and technology. He has a degree in Martian geology and is excellent at playing bagpipes.