Episode 13: Robotics and the COVID-19 Response
In this podcast, Dr. Jason Moats, Director of TEEX Testing and Innovation Center, moderates a panel of experts from Texas A&M University who participating in a National Science Foundation sponsored project examining how robots were used in response to COVID-19. The panel discusses the research as they affect emergency managers, public health providers, and those who have had to converge to create operational tools during the last twelve months.
Speakers include: Dr. Jason Moats, moderator (TEEX), Dr. Robin Murphy (Texas A&M University), Dr. Angela Clendenin (Texas A&M University), Vignesh Gandudi (graduate student, Texas A&M University), Trisha Amin (undergraduate student, Texas A&M University)
Visit https://roboticsforinfectiousdiseases.org/ for more information.
Narrator: Welcome to the Disaster Management Podcast series. Each episode features subject matter professionals discussing strategies and techniques for emergency managers and policy makers to consider as they prepare for, respond to, and recover from disasters. This series is brought to you by Texas A&M Engineering Extension Service.
Jason Moats: Howdy and welcome to our podcast on robotics and the COVID-19 response. My name is Jason Moats. I’m the director of the TEEX Testing and Innovation Center here at Texas A&M Engineering Extension Service. Today on our panel, we have Dr. Robin Murphy, who is a professor of Computer Science and Engineering, Dr. Angela Clendenin, who’s an instructional assistant professor of Epidemiology at the Texas A&M School of Public Health. Vignesh Gandudi. He’s a graduate student in Computer Science and Trisha Amin, who’s an undergraduate student at the School of Public Health. Folks. Welcome. I’m delighted to have a panel of folks that I’ve been able to collaborate with over the last few months looking at how robots are being used in response to COVID-19. We’re going to get into panel introductions here in just a moment. But this is going to be an in-depth panel where we are discussing issues that affect not only emergency managers and public health providers but also folks that have to converge together to create operational tools. I’m going to ask each of our panel members to introduce themselves and with a little bit of background on their relationship to the project. And we’ll start right now with Dr. Robin Murphy.
Dr. Robin Murphy: Well, howdy, everyone, I’m Robin Murphy, I’m a professor of Computer Science and Engineering at Texas A&M and I work in disaster robotics. And in particular, I’ve been documenting and analyzing how robots have been used for pandemics since the 2014 Ebola outbreak. This COVID outbreak, of course, is a big moment for us.
Jason Moats: Thanks. Next, I have Dr. Angela Clendenin.
Dr. Angela Clendenin: Hello, everyone. I’m Dr. Angela Clendenin. I’m a disaster epidemiologist and instructional assistant professor at the Texas A&M School of Public Health. I’ve got about 10 years of emergency management and deployment experience with a veterinary emergency team. And I’ve been interested in this novel illness and how we’ve been able to use automated systems to respond better.
Jason Moats: Thanks, Dr. Clendenin. And Vignesh Gandudi.
Vignesh Gandudi: Hello, my name’s Vignesh Gandudi. I’m a graduate student pursuing my research under the supervision of Dr. Murphy in the Department of Computer Science at Texas A&M. What interests me in this project is the adoption of robots for COVID-19. And I was involved in collecting the data through various data sources and filtering them to see the explicit use of robots for COVID-19
Jason Moats: Thank you. Last but certainly not least, Trisha Amin
Trisha Amin: Hi, I’m Trisha Amin, I’m an undergraduate student at the School of Public Health at Texas A&M. My primary contribution to this project, was creating graphical representations based on the data we collected.
Jason Moats: As you can see, we have a great panel full of experts. And I would be remiss if I didn’t mention this project was funded through the National Science Foundation. And as we get started into the panel, one of the things I want to ask is, Dr. Murphy, this is for you. Why this project, why now, what’s the motivation behind it?
Dr. Robin Murphy: The motivation from the project, there’s a couple of motivations of why we’ve spent a year collecting everything we can about how robots have been used. And one of the motivations is that it’s just good to document this. This is going to help roboticists and practitioners both understand what robots can do. But another reason is that most disasters are limited in terms of time. They’re over in a couple of weeks. They’re limited to a single country or state. So we never really get a lot of data about who’s using what robots, why, and if they’re working out. So it is a unique opportunity to learn both about robots for pandemics since almost every country has been involved with it as well as overtime. And then to see how much of that will transfer to us for just regular disasters. I mean, it’s getting to where a wildfire, a flood, or hurricanes is looking good compared to a pandemic.
Jason Moats: Absolutely. This is an amazing opportunity to be able to look at this almost in real-time. That kind of leads into the next part. Dr. Murphy, can you kind of give us an overview of this project and talk about what we’re trying to achieve through what we did.
Dr. Robin Murphy: To give you an overview of the project and the data is that collected over four hundred reports. These are press releases, articles on the news, scientific literature, articles, posting on tweets, and going through and seeing if there were robots. You know, a lot of times just the word robot and COVID shows up in pretty much every language. So keying off of that. And out of those four hundred reports, there were three hundred and thirty-eight instances of where robots were being used explicitly for or because of COVID. And that was across forty-eight countries. I mean, just an amazing amount of robots, more than we expected. And of those, the majority were ground robots, but a third were drones. And then there were a couple of marine vehicles, robot boats as well, to help out.
Jason Moats: Ok, after doing all this analysis, what were the findings. I mean, how does this relate to emergency management and public health, and robotics?
Dr. Robin Murphy: Well, I think the first finding, I think, for emergency managers to think about is that there were a lot of robots that were used, ground robots, aerial robots, and even those marine robots. Start thinking about the way different robots can be used. I think a lot of people just thought of robots, as soon as they heard of robots, as it must be somebody trying to replace a doctor. We certainly had a couple of those experimental robots to do the nose and mouth swabs, but the majority weren’t and it was a very broad use and that would be the second big finding. Robots have been used for almost everything. We found six major categories of how robots were being used. There were, of course, things that we’ve called clinical care. We’re taking care of disinfecting the hospitals about telemedicine. But you’re also having delivery robots helping and patient and family socializing. Clinical care seems the most obvious, but it turned out that was the second-largest number of reports. The largest number of reports was used by public safety and public health groups for things like quarantine enforcement and also taking these agricultural sprayers that these drones are doing precision agriculture, taking those sprayers and disinfecting public spaces.
More surprising to me is the other things they get used for. They got used for continuity of work and education. People started using these robots. Private companies started using robots in ways they’ve never used before. Universities and educational facilities started using them. Not a big surprise in the laboratory and the automation and supply chain, those beefed up what they did and individuals started using robots in all sorts of different ways. Certainly, you had a consumer base like where a lot of delivery of food and non-food purchases were done. But you also started seeing people very creatively using drones to attend public events and robots to attend graduation and ask people out on dates and all sort of walk a dog. I mean, all sorts of creative ideas. And then the smallest case we saw and we didn’t see it much of the United States, but we did see in other nations was known hospital care using robots to help take care of people in a nursing home that were being quarantined or delivered to them, monitor that or if they were in one of the quarantine camps or hotels.
Jason Moats: That’s a lot of information to come out of the study. I want to ask the panel, we’ve all been intimately involved in this and I’m curious, and we’ll start with Dr. Clendenin, were there any big surprises to you about what we found as we looked into this?
Dr. Angela Clendenin: I think one of the most interesting findings or surprising findings for me is with this novel disease and trying to look at the disease trajectory within an individual who contracted COVID and having to be isolated for 10 days in quarantine for 14 days for close contacts. That’s much longer than most infectious diseases we’ve had to deal with on this scale and to be able to see how people incorporated robots and automated systems to try and have some sense of normalcy in their lives in terms of socialization, access to services, even the diagnostic end. I think it was really surprising. It’s like the pressure that was applied by this pandemic fostered this much broader adoption of innovation in the lay public outside of just the roboticists community. I was intrigued by this, the breadth of views that we saw.
Jason Moats: Vignesh?
Vignesh Gandudi: For me, I would say the most surprising thing would be from an adoption perspective. That’s what I found interesting, is that most of the time we do expect that we adopt the robots which are already existing, which was the majority in our case. There was 78 percent of robots which are already heritage, which was categorized according to the theory in other categories. But in spite of that, we had it on thirty-nine instances of new robots where most of the robots, the new robots were not adopted in first world countries. They were adopted in Third World countries like India and many of the African nations. I found that surprising that new robots are being adopted in different countries.
Jason Moats: Tricia, any big surprises
Trisha Amin: Similarly to Dr. Clendenin’s statements. It was interesting to learn about the variety of uses through the different word domains. And I also found it interesting seeing the interactions between or the implications for ethics and policy when it comes to such a wide use of robotics, especially since this is sort of unprecedented. It’s new territory and seeing how humans and robots interact in different spaces.
Jason Moats: Dr. Murphy, I certainly want to hear from you. You’ve listed you listed a bunch of surprises, as you were looking at. But was there one that just really stuck out to you that kind of left you going, “I didn’t expect that?”
Dr. Robin Murphy: One of the ones that took me by surprise was the idea that robots would be used in so many different ways. There are six different work domains. When we started looking at the data from an international level and you saw the same pattern. It was the countries that used robots for more than one thing, more than disinfecting a hospital. They started almost immediately to both start using robots for clinical…start using robots for public safety…start using robots for delivery to homes and people. It was amazing that it was like the creativity was just incredible so that if you were going to use robots, you were going to think about them in multiple ways instead of I just thought it be, you know, clinical care and maybe public safety, maybe some lab automation. But no, it was much broader for the top countries that use the robots. What makes it even more fun to think about is that those robots that are being used, we found twenty-nine different use cases or applications. Twenty-seven of those twenty-nine applications already had robots that were commercially available and being used in that application. Now, they may not have been in that particular country or they may not have been in that hospital but they had those. The idea that you had this big base of robots to draw off of and could use, I think was something we’ve never seen before in robotics, much less disaster robotics.
Jason Moats: There are two questions based on something that Trisha said and something you just said, Dr. Murphy, which is I want to go to the first one: Which is did the data kind of give you a breakdown based off of the country? How did the use of robots vary by country?
Trisha Amin: When we looked at the uses by countries, we were able to distinguish the top eight performing countries in terms of the number of instances that they had. One thing that was interesting in my opinion, was that the United States had instances in every domain except for non-hospital care while the rest of the countries had a more equal distribution amongst the different domains.
Jason Moats: This is for the whole panel, is this a function of national policy? Is it a function of innovation or is it just being in the midst of a crisis? How does policy play into this?
Vignesh Gandudi: What we observed was out of the top eight performing countries or the top eight countries which have the greatest number of instances. We had six countries that already had the national level or robotics initiative already present. I would say, yes, the pre-existing policies are of pre-existing robotic initiatives actually complemented the adoption.
Jason Moats: The other question that has popped into my mind is we’ve talked about these positive uses and functional uses of these to deal with the pandemic from many different spectrums. But my question is, did we find any indication that robots were being used inappropriately in response to COVID-19?
Dr. Robin Murphy: We have found some indications that robots are being used inappropriately. We’re still looking and analyzing that data. But as of the July data set point, we found forty-five different reports of concerns about them. A lot of these are things the robots are being used for unwarranted surveillance, violating privacy, violating intellectual property. Some of the concerns are concerns that are a fear that they are displacing workers or a fear that they aren’t regulated enough, that certain things aren’t happening. Or we see in some cases some questionable judgment calls in terms of professional ethics. But out of the four hundred reports we’re getting and having only about forty-five with ethics indicates that it is fairly small that is the first thing. We did see that the majority of those ethical concerns were associated with drones and in particular drones in public safety. And I think that’s a big lesson for the emergency management community, is that the drones, they weren’t violating any aviation regulations, but they were being typically used very aggressively for surveillance, for quarantine enforcement. Depending on what country you’re in, often generated a reaction. And then the other one is that they’re used heavily and some of the Asian countries for disinfection. Spraying lots of disinfectant over a subway, a public park, streets, things like that. Well, two things immediately came up. One is that that’s a lot of toxic chemicals that you’re putting in that has a runoff problem and you’re doing it to an open population where it wasn’t clear you were getting them out of the way. The World Health Organization (WHO) came in and said, time out, let’s stop this. You’ve got to be very careful because you don’t want to cause more health concerns or take out any environmental impacts as well. I think this is something for us to consider. As in the emergency response, field regulations weren’t a barrier, certainly aviation, the health know how to relax those. I think there was a misjudgment on the case of roboticists kind of being that usual hype and enthusiasm. But I think it’s down to us to think through how we’re going to use the technology and the downstream consequences.
Jason Moats: Well, there’s a question I have that goes back to something Vignesh said about heritage robots. Dr. Murphy, can you explain heritage robots and what that means?
Dr. Robin Murphy: Heritage robots is one of three categories or ways of rating the technical maturity of a system. Most people I think, are familiar with NASA’s technical readiness levels of zero to nine. From a great idea and then down to nine, which means it’s in production, standard use. Well, NASA has expanded that to consider things like workflows. A widget can be used a lot but if you use it in a new situation or a new application, it can have some downstream effects and impacts. They started classifying them into three categories, the technology heritage, which meant it was physically mature and the application was well understood and in practice. Engineering meant that you were just changing it a little bit, either the hardware or a very similar application, but kind of different. You might have some unintended consequences or things to watch with your workflows and your manpower allocations. And then the third category was new, which meant that you either had a very novel application, something quite different, or you’re taking a new robot or platform or combination that you’ve never done before. That gives you a much better feel for not only is the technology like out of a lab and it works and it won’t physically fall apart, but whether or not it’s been used by someone like you with this particular application before and so it’s very trustworthy.
Jason Moats: Wow. That’s a different way of thinking about it. It’s more like the mission readiness as opposed to something that could be at a high TRL level, but not ready for the mission. Interesting. Another question that’s come up has been how this has unfolded. Robot use unfolded with the pandemic, and I’m sure that you all looked at that. Can you talk about how we saw the use of robots unfolding as it correlates to the epi curves that we’re all familiar with?
Trisha Amin: We also compared the reported uses of each topic country with their respective epi curves. The epi curves, in this case, were the daily case counts of COVID in that country. And an interesting pattern that we sort of observed was that the countries with policy initiatives showed reported use either start before or around the same time as their initial uptick in cases. One exception to this was Italy. It was interesting to see that their reported uses picked up after their initial uptake in cases rather than before or during compared to the other countries. Something else that we noticed was that the reported use for all of the work domains picked up around the same time and then eventually leveled off as the disease progressed in the country.
Jason Moats: Vignesh had talked a little bit about the heritage robots. I’m curious, did you see new robots? Did you see innovation happening? Can you talk a little bit about how that unfolded?
Vignesh Gandudi: With new robots what we saw was new hardware, software, and there was something like new missions that was like, for example, maybe the robot was already existing, but you were applying it for a completely new mission, which you haven’t done before. For example, identification of infected. We had infrared cameras before, but fixing them on drones and using them to identify the infected in public spaces, that wasn’t done before. That was something like a new mission that we saw during the pandemic. Even though we can’t guarantee the efficacy of it, whether it’s effective, identifying the infected. We saw some in many cases of this was something we speak of new missions and there were some cases of robots where they built the robot for a particular use case or let’s say mouth swabbing. They built new hardware to automatically do the mouth swabs and nose swabs. These were some of the instances that we saw for new robots.
Jason Moats: I want to go around and ask about what the implications are. Dr. Murphy, what are the implications for the roboticists? What are the things that roboticists should look at as we go to our next big event, whatever that may be.
Dr. Robin Murphy: I think the big telling lessons for the robotics community is that nothing replaces working with the end-users, with the emergency managers, with the responders before a disaster. Understanding those use cases because the majority of use cases already existed. If there was a low number of really new ideas or new leads that emerge. The same thing on the robots, the robots existed. What is it that we need to be doing? We need to be adapting and making the robots easier to be used in the workflows, easier to adapt for slightly different situations, slightly different environments, slightly different missions. It’s less about, oh, pretty shiny new, it’s more about sitting down and doing the work to understand the system.
Jason Moats: I know from an emergency management perspective, which is one of my main domains, what struck me was that there were many heritage robots used throughout. Which means that they existed and I wonder how many emergency managers, emergency response organizations understood that before COVID. How were these being integrated? We’ve talked about them for surveillance, for doing assessments, and things like that. We’ve seen that. I mean, here in Texas, we see those with hurricanes and floods and things like that. We’ve seen those missions, in fact Dr. Murphy’s even flown those missions. But some of these other pieces have been phenomenal, and I think that it’s one of those things that from an emergency management perspective, where we need to engage with the roboticists and say, here are the problems we have. How can we use robotics to address, mitigate, solve these problems? How can we have it so when that next one occurs, we can deal with it?
Dr. Robin Murphy: I would like to follow up on that because I think that what this mostly says about disasters is that we’re now at the point where what we have as a community is good enough and that responders should they may not have been using it themselves, but their colleagues somewhere have. There are options out there for them. They just may need to reach out. That regulations are a problem. We’ve got some solid stuff and I hope they’ll be able to use them in the future. We roboticist, will do better and keep making more but working with you.
Jason Moats: So, Dr. Clendenin, a similar question to you as we’ve gone through this research and we’ve looked at the different cases, I’m wondering from a public health and emergency management perspective, what do you see being the major implications? What is the major take-homes for those communities?
Dr. Angela Clendenin: I think any time you have a response like this, whether it’s evacuating from a hurricane or quarantining in a pandemic, the number one challenge that emergency managers and now, in a pandemic, public health professionals have is compliance with guidance. And, you know, it’s really hard to get people to stay home for 14 days or to isolate for 10 days or to disconnect from what they know, which is their solid foundation. But when you can see the innovation that occurred because of the COVID-19 pandemic that answered some of these emerging needs of individuals to address their psychosocial and their work lives, I think that’s the big implication out of this. When we see these kinds of demands, it fosters the innovation to respond to these not just economic safety needs, but also from a public health perspective. Being able to address the psychosocial needs of a community helps improve compliance with guidance. And if we’re ever going to get on the other side of this thing, we have to be able to improve compliance. If we can look for a technological solution, which we appear to have been very successful at in the past year, I think we can see some improvement in compliance in different types of disasters moving forward.
Jason Moats: Trisha, you’ve been poring through all this as well, and you sit in a very interesting position as an undergraduate student. I’m curious to know, what do you see as those implications for both the roboticists and those working in public health?
Trisha Amin: I think this pandemic has given us a good opportunity for those two groups to come together and see what’s working and what’s not working as pandemic progresses and as we deal with new issues as they come. I think it’s interesting to see how there are newer or ways that we didn’t think of to implement robots in a public health response. Specially going in the future as we face other inevitable disasters.
Jason Moats: Vignesh, same question, what do you see as major implications from this as we move forward?
Vignesh Gandudi: One of the things that I observed was that the kind of robots that the majority of the people used of the general-purpose scenario, that there was a robot, but the same robot was used for multiple applications. I would say from a roboticist’s perspective or a developer perspective, at least for the future of inevitable disasters, we should direct our research towards robots or technology that can be easily adapted with minor tweaks or with minor changes where we can easily adapt if we have a new mission or a new application that we want to apply it to. That should be possible in the future. It should not become a very niche technology that everything has to be done from scratch for that particular application. It should be something that we can go fix the existing one, tweak it a bit, if it’s necessary, and then apply it towards an application. From a robotics perspective, I think that would be useful.
Jason Moats: Excellent. We’ve just about reached our time. I’d like to go around and ask the group; do you have any last thoughts? What is your advice? Moving beyond COVID-19?
Dr. Angela Clendenin: As a practitioner of public health, one of the things that we had to try and make is our mantra moving forward through the pandemic from the very earliest days in trying to get people to do what they were supposed to do was, don’t focus on what you’re being told that you can’t do, focus on what you can do. Sometimes we can still do the things that we were wanting to do every single day. We just have to be a little more creative about how we engage in these activities to keep everybody safe. I think this research points to a particularly fruitful relationship moving forward because to be able to be creative and do things in a pandemic, we need to be engaging some different disciplines, including robotics and emergency management, public health, all coming together. I think that this project shows the power of a truly interdisciplinary approach to pandemic response.
Jason Moats: Dr. Murphy.
Dr. Robin Murphy: One of the things that I would hope would come out of the COVID pandemic that we didn’t see as much as we would have liked out of the Ebola pandemic is continuity of effort. We saw a lot of interest in robots during Ebola and then it just went way down, two weeks after, way down, just down. So I think if the sheer number and value of the robots that have been used if we can keep the government incentives, the insurers allow these to be used so that as soon as the pandemic finally passes, that these will stay in use these robots, and then they’ll get more reasons for companies to invest, to do better, to work more closely with public health and public safety, to build the next generation so that when the next inevitable pandemic and other disasters come, we’ll be in better shape.
Vignesh Gandudi: I agree with what Dr. Murphy said. What I would want to see is more acceptance of robots going forward. We have been using robots in pandemic applications and more complementing the existing scenario. Most of the applications that you saw, the robots were complementing. They were not taking away something that was already existing they were complimenting it. They were making up for the deficit scenario. Let’s say, a disinfecting hospital already a huge volume, but the robots came in as the rescue for that to balance that deficient scenario. What I would want to see is that going forward, pandemic or no pandemic, the acceptance of robots where people utilize the scenario, where it complements the existing working applications.
Trisha Amin: One thing that I would like to see is that. Obviously, at this point COVID is probably the most salient public health issue that we’re dealing with, and because of that, I think some other concurrent issues have been placed on the back burner. But I think it’d be really interesting to see any of the discoveries or developments that have occurred during or because of COVID. I think it’d be great to see them applied to current public health issues such as the influenza epidemics or pandemics that we see on a seasonal basis. I think it’d be interesting to see how we could apply what we learned in COVID to something that we’re already dealing with.
Jason Moats: Excellent points by everyone. Well, our time has come and it has been a pleasure being with you all this afternoon. What a fascinating project. Working and looking at robots as they’re being used in response to the global COVID-19 pandemic as we conclude this session. It’s been my pleasure to be with our panelists today, Dr. Robin Murphy from Texas A&M and Dr. Angela Clendenin also from Texas A&M. Vignesh Gandudi, a graduate student here at Texas A&M, and also Trisha Amin, who is an undergraduate student. We’ve all been working together on an interesting and insightful research project, looking at how robots are being used in response to the current COVID-19 pandemic. My name is Jason Moats and until next time, thanks, and have a great day.
Narrator: Thank you for listening to the Disaster Management Podcast series, brought to you by Texas A&M Engineering Extension Service. Visit www.teex.org for information about training near you.