Main Engine Cut Off - T+227: Astrobee, Free-Flying Robots on the ISS (with Jonathan Barlow)

Episode Date: August 22, 2022

Jonathan Barlow from the NASA Ames Research Center joins me to talk about Astrobee—the free-flying robots on the International Space Station that test new technologies, sensors, and operations, and assist astronauts and mission controllers alike. Jonathan is the Astrobee Lead Engineer and has been with the program since the beginning, so we go way back to its origins and talk in-depth about what the robots are up to today on the ISS.This episode of Main Engine Cut Off is brought to you by 43 executive producers—Simon, Lauren, Kris, Pat, Matt, Jorge, Ryan, Donald, Lee, Chris, Warren, Bob, Russell, Moritz, Joel, Jan, David, Joonas, Robb, Tim Dodd (the Everyday Astronaut!), Frank, Julian and Lars from Agile Space, Matt, The Astrogators at SEE, Chris, Aegis Trade Law, Fred, Hemant, Dawn Aerospace, Andrew, Harrison, Benjamin, SmallSpark Space Systems, Schultzy, and seven anonymous—and 818 other supporters.TopicsWhat is Astrobee? | NASAAstrobee - Four Years in Video | NASA Image and Video LibraryNASA Celebrates Three Years of Astrobees Buzzing on Space Station | NASAMeet ISAAC, Integrating Robots with the Space Stations of the Future | NASAAstrobee Space-Bots Mark a New Milestone in Human-Robot Teamwork | NASANPS, NASA Team Up on “Astrobatics” Project to Advance Spacecraft Robotics - Naval Postgraduate SchoolArtemis 1 Launch Events - Off-NominalThe ShowLike the show? Support the show!Email your thoughts, comments, and questions to [e-mail] @WeHaveMECOListen to MECO HeadlinesJoin the Off-Nominal DiscordSubscribe on Apple Podcasts, Overcast, Pocket Casts, Spotify, Google Play, Stitcher, TuneIn or elsewhereSubscribe to the Main Engine Cut Off NewsletterMusic by Max JustusArtwork photo by ESA

Hello and welcome to Main Engine Cutoff, I am Anthony Colangelo, and I've got a fun one today. We are going to be talking with Jonathan Barlow, who is the Astro B lead engineer. He is out of the NASA Ames Research Center. This one has been in my mind for a while because if you're someone who listens to MECO Headlines, the show that I do every single week for supporters of the show, we go through all the stories of the week, but we also talk about what's going on up on the ISS each and every week. And a lot of times, Astrobees are mentioned. These are

three free-flying robots that operate on the space station. They are testing out different sensors and payloads and technologies, different operations, trying to explore the ways that robots can be helpful alongside humans within the ISS. Some of the recent things that have been coming up on the show is this research that they're doing to use audio from the space station to detect anomalies before they actually result in something catastrophic. So the astrobys fly around the station, they listen to the general audio of the space station, all the sounds on board. And if something sounds different, then they're able to know, okay, something's either not operating or operating differently, or, you know, there's something more to look into in this area. There's just tests like that going on all the time.

So you're constantly hearing about an astronaut setting up an Astrobee for some sort of test or another. And it just has always interested me a lot. So I reached out and Jonathan was happy to join me to talk in depth about what Astrobees are doing on the station. But we've also, we'll dive into the history because he's been with the program for a very, very long time. So we talk about its origins, the programs that he's been with the program for a very, very long time. So we talk about its origins, the programs that came before it, how Astrobees are different, how they fly around the space station.

It's a really fun conversation about a really interesting technology that I feel like a lot of people don't know about. And, you know, I've certainly wanted to learn more. So I'm excited to talk to Jonathan. So without further ado, let's give Jonathan a call. All right, Jonathan, welcome to Main Engine Cutoff. Thanks for hanging out with me for a bit to talk about Astro Bees. Yeah, thanks for having me. So to start, you are the lead engineer, is what I was told. Is that your official title on the project? That's what I do. Yes, I'm the lead engineer for Astro Bee.

So before we get into the technical details, let's start with some of the background about Astro Bees generally, but also how you got involved with the project and how your timeline has gone alongside Astrobees. So my background with Astrobee, I actually was fortunate enough to be on the team of people that was putting together the proposal to propose to create Astrobee. I was working for a project called Spheres, which was a series of three satellites that were on the space station that were doing research and people would come and do research with them. And there was this idea of, hey, could we do something better, something more, you know, renewable? Spheres had one-time use

batteries and CO2 cartridges. So we thought, hey, wouldn't it be cool if we could have rechargeable batteries and maybe not have to replace CO2 every time? So we worked with the team at Ames to come up with the proposal and we got it funded. So I've been with it ever since. So I've been with it ever since. I worked on the development of AstroV as the deputy systems engineer and the integration and test lead. And then when the development finished, the AstroVs were turned over to what was then the SPHERES facility and is now the AstroV facility. And we started using that. So I'm the lead engineer for that as well.

In terms of how the program got started, you mentioned SPHERES. From what I've seen, those were a lot smaller than the Astrobees that are flying on station today. Was that also part of the expansion of the idea of what you could do with these kinds of robots? Or was it driven by what you just mentioned, like trying to limit consumables and make them a little bit easier to manage? What was the driving force behind size, scale, scope, and all that? Yeah, so size is largely driven by how much mass we had. So we wanted to be able to run for a long time. With our rechargeable batteries, you know, we can go back and charge them, but it's better to be able to run for a while. So that kind of drove up the mass of the batteries, kind of drove up the mass of the overall system.

And then the size is dictated by how quiet we want it to be. We're fan-based propulsion, and so the larger the fan you have, the quieter it is. So we wanted to meet certain requirements, limits on how long we could operate based on the noise that we were generating so we were trying to keep under a certain threshold let's talk about the free-flying nature because when i've talked to some people about astrobys in the past yeah so i do this other part of the podcast where i keep everyone up to date on headlines every week and part of that is reading through the iss blogs talking about what's been going on in the station and every time astrobys would come up somebody'd be like can you talk more about those because that sounds really cool um and a lot of times when i start talking about that there's this moment of like oh yeah i never really considered

how in zero g you could just use fans to fly around autonomously like that uh so it's how does that fan system work within this what is the you know is there just there i think there's two fans with internally but how does the the actual propulsion operate in that environment? Yeah, I wish that I could point to a diagram. On the sides of Asterbee, there's two propulsion modules, and they're basically copies of each other. In the center of the module is an intake where the air gets drawn in by an impeller fan, that does a very light pressurization of the rest of the module. And then there's some nozzles that are distributed around the propulsion module to be able to let that air out in a certain direction and create reaction forces.

Okay, so I think I have a diagram that I'm looking at from somewhere on the site. And there are nozzles on each of the axes, or are they controllable? Yep. So it's what we call holonomic. So it can rotate or translate in any direction from any given attitude. There was also this really cool ground prototype test harness

I saw in some of the videos. I'm curious how you developed this on the ground um it's not really something that's easy to create an analog of something that can fly around and you know six degrees of freedom in a space station so um that process to create these ground prototypes what did that entail how did you come up with the ideas of how to actually rig these up on the ground to make sure that the testing program went well before it went up to the station? Yeah, so there's kind of two questions there. One is all the prototypes that we did. We did a lot of prototypes, and a lot of it was to test this new propulsion system, because we didn't know

what it was going to be like. We needed to meet our noise targets and that sort of thing. So there's one video on the website that shows like three or four different uh different obviously different prototypes and so we went through a lot of iterations trying to to narrow in on what we were going to do and how we were going to do it but then the something that i think was is also relevant to that question is the the test setup in the in our lab which we have a bigite table, which is, you can think of it as a reverse air hockey table. The air comes out of the puck and it floats around on the table and it gives us a 2D frictionless surface. Well, practically frictionless. And we put up some visuals that

mimicked the ISS environment so that the vision system could have something to look at that was relevant. And so we did all our testing in that environment and it could have something to look at that was relevant. And so we did all our testing in that environment. And it was really great to have that available to be able to actually test because we can put an actual unit on that puck and float it around and have it do its own thing. I'm curious to figure out exactly when you were... I guess maybe we should talk a little bit about the structure of the project overall, because I see a lot of mentions that if you're a scientist that wants to put your own payload on an Astrobee to do work on the station, that's something that the platform is meant to support. So could you talk a little bit about the division between the platform itself, the users of it, and how that interacts with the ISS program? Yeah, so Astrobee is a cool thing. Like you said, people notice it, and how that interacts with the ISS program?

Yeah, so Azure B is a cool thing. Like you said, people notice it, they want to know more about it. I love talking about it. It's a robot, and on its own, it's pretty cool. But some of the science that we are able to do is actually much cooler. So we support lots of different universities, companies, international partners, our own internal NASA projects, those sorts of things. We got lots of projects. Some of the ones that I might talk about are Realm, which was an RFID reader system. They have hatch readers and things on station that are tracking bags and things. It's automated logistics because ISS is a big thing. There's lots of stuff and it's all floating around and we want to make sure we know where things are so it doesn't take

astronauts longer to find things. So one of the things that they needed was something to go and look for lost items. They had the hatch readers, they had readers and drawers and all these stationary readers. But if something's not showing up on a reader, where, how do you find it? So they came to us and they built a little piece that plugs into Astrobee that has an RFID reader in it. And then we're able to move that around and they're able to look for lost items. So that's, that's an example of one. And that's a NASA project. Soundsea was one that was a Bosch, an astrobotic payload. And they were trying to map the sound environment inside the International Space Station. So they had a multidirectional microphone that attached to AstroVe. And we moved around and they were gathering sound data, trying to create like a heat map of loud places and quiet places, those sorts of things.

So in the design phase, how did you determine which sensors and which components needed to be part of the core platform versus things that you would let others be added on in the future? What were the decisions behind that? What did and did not get included? in the future? What, what were the decision behind that? Uh, what, what did and did not get included? Yeah, well, there were things that we wanted to include that kind of got cut. Um, so, um, a lot of it was again, size and weight. Um, uh, they, we focused on three main jobs for AstroV. One was as a mobile sensor platform. So putting a sensor on there and moving it around. One as a mobile camera. So if a ground operator needed to see something or wanted to have eyes on something and didn't have crew time to be able to do that, we could fly over and put a camera on it. And then as we've already talked about, the research goal of being able to host

payloads and have them do those sorts of things. So there was a lot of things that we tried to include, thinking that researchers might want to use them. We have some human-robot interaction type things, LEDs, lights, the touchscreen, you know, with eyes on it on it i mean there's a lot that could be done with those that are just kind of um there and yeah i think a lot of it had to do with our cost and schedule we had to we had to cut out stuff when when we were coming up to the end of what we could do let's talk a bit about how it finds its way around the station um there's a couple different cameras on board a couple different uh i on board, a couple of different, I saw like there's HasCam and PerchCam and DockCam and NavCam.

So there's a bunch of, you know, vision and LiDAR based stuff going on there. How does it actually make use of that when the station is an ever-changing environment? How does it find its way around? It's interesting you should mention an ever-changing environment. How does it find its way around? It's interesting you should mention an ever-changing environment. So when the vision system was first envisioned, we thought, well, it'll be static, relatively static for like a week maybe. So we learned from experience that

actually it changes a lot more than we realized. It's a messy place up there, you know? It is a messy and a busy place. There's a lot going on. They're doing great stuff. But yeah, all the different cameras. There's actually the NavCam. It's the navigation camera. It's the main camera that's used for navigation.

The HazCam is facing forward and is intended to look for obstacles and stop the astrobot if there's an astronaut or something in the way that wasn't expected. Then on the back, there's a dock cam, which we use when we're docking because we have to back in to the charging berth. And the perch cam was intended to enable us to detect and perch on handrails. So it's a 3D. So the HA haz cam and the

perch cam are both the same 3d depth camera sensors and the nav cam and dot camera very similar slightly different lenses on them um one is intent is used all the time for during navigation and one just when we're docking um there's a smattering of other there's the science camera which like i said was the you know the main camera it's a smattering of other, there's the science camera, which like I said, was the, you know, the, the main camera, it's a high quality cell phone glass, high, high def camera. Um, and then there's like a speed camera, which is used as a, as a secondary speed cutoff, uh, for safety. So does it, it uses vision based navigation primarily. And, um, is that refreshed occasionally based on the the changing environment how does that actually implement yeah so the vision-based navigation uses natural landmarks so that means

that we use the astrope we fly it around and capture a series of images and then we build those images into a map and that map we generate offline. And that map we upload to the AstroVean that it uses to localize that. We do refresh that periodically. We've done some localization system updates, which help us to smooth over the gaps for when we're not seeing features in that map. And those have been really good in making

AstroVean a much more reliable platform but those those maps I think we refresh now I want to say quarterly or every other month we have a regular activity that we do that's not associated directly with a research project that is just for you know maintaining the system and making sure we have the maps up to date and we do tend to use that for some secondary objectives to test little bits of other people's things just to make sure that we have everything running but it's kind of a facility focused activity that we do what is the breakdown between operations and research projects that you mentioned? Are there things that are, you know, standard, like the AstroBees get used for these

kind of ISS operations today? Or is it really like AstroBee maintenance and then research projects? So AstroBee's main goal, main job so far has been research, you know, we enable research for autonomous robotics, and we support things like the moon to Mars and Gateway and those sorts of things. We're enabling those kinds of research. When I say we support those, we're enabling the kinds of research that could go into that. So that's his main job to date. There have been a lot of discussions in the past few months about what other kinds of jobs that AstroVee can do. We were now starting to get, because of the reliability that we've built up with AstroVee, we're starting to get people in the ISS program asking, hey, can AstroVee, you know,

inspect these hatch seals for us? Or can AstroVee inspect this? Or can AstroVee build a panorama of, spec this or can AstroPay build a panorama of, you know, this, this module, that kind of thing. So I'm looking forward to, to seeing what other jobs AstroPay might get that aren't specifically research focused and that are kind of more of that mainline ISS support type thing. Yeah. I don't know if it was the SoundSee program. The one that was mentioned a lot in the last couple of months was using sounds to detect anomalies or things not running correctly on the station. Is that the same project? Yeah,

that was the sound C mapping the ISS environment and detecting anomalies. Yep. Yeah. Cause that one sounded like when I started reading that, I was like, Oh, that's a brilliant kind of like passive monitoring of a system that was, you know,

designed to be operated by humans and here we are in 2022 it's still flying up there so things have changed um it's interesting to consider that kind of thing uh i don't know maybe i'll broaden the question a bit like the iss was designed in a different era of spaceflight it was you know decades ago that the i mean if you roll the clock back far enough, it was, you know, space station freedom back in the eighties or whatever it was that designed some of the initial ISS concepts. Are there things that are specific about the environment that ISS is, which is a human tended, constant habitation space station that lends its hand to the way that astrobys operate? Um, you know, I guess I'm comparing

this against some of the concepts that are out there of like a robotic space station that does manufacturing and occasionally humans come up and collect stuff like i don't know there are differences that i could foresee in the way that stations are built do you feel like astrobys lean into the human tended station concept in any particular way well yeah astrobys was designed with the human tended station in mind and we certainly take advantage of that um when we do get lost and or get hung up on a cable or something we do occasionally get a crew member floating by saying hey you want me to put that back somewhere for

you it's like when your robo gets stuck under a table you're like oh i gotta get this guy moving again you know all robots and even you know your little rc car will do that but um so we we definitely take advantage of that and that's what we were designed for but like i said we're we are helping to enable the kinds of research you're not going to have a fully autonomous station unless you have done the research and figured out how to run a station autonomously. So I think that would be a marvelous thing to see in our time, a station that is mostly robotic. And there's people talking about doing it, and I hope they are successful. I mean, even NASA is to some extent, right?

Gateway is only going to be crewed occasionally, basically, on the current mindset. So I have to imagine this is a big part of of uh what gateway would take advantage of so some of that passive monitoring um i know there is there's some sort of arm on this i don't know it might be related to the handrail perching that you talked about um are there you know concepts about what if we gave this thing a couple arms and it was a little uh swiss army knife that flies around the station and it can do sorts of stuff that that human hands could also do oh wouldn't that be marvelous if we had arms that were small enough and light enough to be on that trivia and and do that um yeah i mean there's a there's kind of two ways of approaching how do you do things robotically one is can you

make something that does something like a human and attach it to the station or move it around? That sounds an awful lot like Robonaut. Robonaut was one of those projects that was really cool and had human dexterous hands. I've seen other projects come along that have demonstrated other similar things.

That would be really cool. As a free-floating satellite inside the station, it's hard to grip and provide the forces that a stationary autonomous robot could do. One day, maybe. Maybe someday. One thing I'm curious about is that the, so the, the project came out of, uh, I guess

it came out of NASA Ames originally. That's where this was born and raised. Really. They were built at Ames from what I understand. Is that correct? Yeah, they were. So for programs gateway, even if, if some of the new commercial space station partners are interested in, in having Astro bees on board, what would the process be to build out another Astro Bee or another set of

Astro Bees? Did you keep the paperwork around? Is this something you could build again? Would you transition it to industry? How does that work? Yeah, some of our users have asked for that. They want to build a unit in their lab or they want to replicate certain aspects of it in order to be able to test better. So we do have a publicly released data set that is available for those researchers so that they can do that if that's what they want to do. Yes, we kept all the paperwork around. We still have some of the units.

So when we built AstroV, we built three units for space and three units for testing on the ground and spare parts. So we have still spare parts for what's on the ground so we can do repairs and things like that if needed. Every good space program includes things that you've built not to go to space and spare parts that eventually go to go to space somehow so that's definitely definitely of the oak um last thing i want to round out on is is the operations of this from you know you have a ground segment for sure control um where do where do they get controlled from how do they get integrated to

the space station's plan day to day or research plan overall well that's a great question um sometimes it gets controlled from my bedroom um so let me let me clarify that so we have a place at Ames where we would go for like our group meetings for for doing our operations on on space station and COVID happened, and we weren't allowed to go in. We had designed Astrobee from the start to be able to take commands from our users. We initially had planned to be much more in the background. That hasn't happened quite as much because some of the development that we've been doing to improve our localization

and make things a little bit more reliable. So a lot of our users rely on us to operate it. But because we had built that capability in from the start, when COVID happened, we were able to transition. And a lot of our sessions are done remotely now. We're not co-located. We've had to figure out how to use some of the networking and screen sharing and team communication softwares that everyone's familiar with now that they've gone through COVID.

And so we had to do the same thing, figure out how to get our work job done while physically separated. And that was certainly a learning experience, but we've been very successful with that. It's been great. Do you get feedback from astronauts after they work with Astrobees? Is there stuff that gets incorporated into the operations or, you know, you may keep notes on design for future Astrobees? What is that feedback cycle like? We always get feedback from crew members when they come back. They have like an exit interview and talk about their experience. And we're able to submit questions and ask them, you know, what was your experience?

What do you like about it? What didn't you like about it? So we've gotten historically that information for all the crew members that have worked with AstrobV, which there's going to be quite a few nowadays, they all like it. It tends to be something different, just like we find it interesting on the ground. We get to look at it and learn about it. I think that crew members enjoy looking and watching AstroV do its thing. Sometimes they're just silent observers. Sometimes they're more interactive.

We do some student programming competitions, Zero Robotics and Keebo RPC. So Zero Robotics is a U.S.-based student programming competition for middle school and high school students. And then Keebo RPC is a competition that JA that Jackson runs that uses Astro B and they have, they're in the international competition. Um, well, I guess both of them are international competitions, but, um, but they both use Astro B and they've both been very successful and the crew

members are like sports casters for that one. Um, so I think they enjoy it. Um, you know, you'd have to ask them exactly what they think about it, but I think they like it. I like it. I enjoy it. So it's great. What kind of things are you looking forward to for the future of Astro Bees? They've been up there three years now doing their thing. Maybe even longer than that. It might be four years now. I forget the exact date on that, uh, three year thing. Oh no, it was three, right? I got to look. April is three years per this blog post. So it's been quite a while now what are the things in the future that you're looking at excited about well we have up and coming payloads so we have some that we're still working on there's a

payload svgs which is a smartphone video guidance system which uses a smartphone. It actually uses the smartphone processor that we have available for our guest users to localize in a different way using some active LEDs. And so they're continuing their experiments, which was a really cool thing to watch. Astrobatics is one that's ongoing that's uh it's a naval postgraduate school project that is looking at how do you reduce propellant use um and they are using the arm of astrobee to fling astrobee swinging like a monkey from handrail to handrail so they've done some really cool demonstrations they're not done yet So I'm looking forward to seeing how that, where that ends up. There's one that we call multi-resolution scanning,

which is a, is a company that is trying to do a high resolution mapping of inside the space station. So like millimeter precision type mapping and create a 3d model. Isaac is another project that's actually just wrapping up. So that's not future, but it's something similar with the, with the built-in sensors of Astrobee.

They were able to create 3d module models of the inside of the international space station. So that's all been, it's all been great to watch, but we're looking forward to seeing what other things our users come up with. Yeah, we're going to need some video of the swinging around one. That sounds awesome. Also been terrifying at the same time,

if you're somebody floating through the space station, what is happening there? I guess, I mean, that is how some of the astronauts traverse around, so that's definitely something that they're good at. They can have a race, maybe, see who can get to the get to the other end of the space station first that would be that would be fun to see an astrobute race i'm wondering if one of those student programming competitions one of these

days will do a race so far it's been you know one s3 at a time so we'll see awesome well jonathan thank you so much this has been awesome to talk a little bit about astrobys and demystify some of what's going on up there. But if anyone wants to follow along, I've got a bunch of links in the show notes to places on the web to check out what you all are working on. But otherwise, thanks so much for chatting today. Yeah, thanks for your time. It's been great.

Thanks again to Jonathan for coming on the show and everyone else at NASA that helped set up this interview. They're always very helpful in finding the exact expert that you're looking for when you reach out with a completely random request like I did, which was, uh, Astrobees, can I talk to somebody about that? And, uh, they routed me right to Jonathan. So very, very, very big thanks to everyone over there that helped get this together. And thank you to all of you who support the show out there. There are 861 of you supporting the show every single month. I'm so thankful for that i'm starting to see that four digits in the future a thousand supporters can we get there there are 139 of you

that uh need miko headlines in your life i really do think it's something that you should check out if you like this podcast if you like hearing from me you'll hear from me every single week on all the stories in space uh keep you up to date give you my thoughts oftentimes i'm giving you thoughts a week or two ahead of time what you would hear on this show because I'm still kicking it around in my head a little bit. So it's a great way to support the show. Stay up on Space News and keep more of Miko in your life. One thing I do want to mention before we thank the executive producers is that if you're going to be at Artemis One in the Florida area, Jake and I and the Off Nominal community will be present. We are planning a couple of different events. Things are in flux because the launch is in flux,

but right now it sounds like the night before the launch, we'll be doing something and the day of the launch, we'll have a beach party out at Cape Canaveral. So if you want to come to any of that to hang out with us in person, to see others out there that listen to these shows that think like you do and are, from my experience, some of the most fun people to hang out with. So you definitely should check it out. Offnom.com slash events. If you're in the area, go over there, sign up on the email list that is in the Artemis one events page. That'll go to Jake's email list. He will keep

you up to date with exactly when and where things are happening in case stuff changes, weather launch time, whatever. Uh, he'll be keeping you up to date on everything that's going on down in the Florida area. And let me thank the 43 executive producers who produced this episode of Main Engine Cutoff. Thanks to Simon, Lauren, Chris, Pat, Matt, George, Ryan, Donald, Lee, Chris, Warren, Bob, Russell, Moritz, Joel, Jan, David, Eunice, Rob, Tim Dodd, the Everyday Astronaut, Frank, Julian and Lars from Agile Space, Matt, the Astrogators at SCE, Chris, Aegis Trade Law, Fred, Haymonth, Dawn Aerospace, Andrew Harrison, Benjamin, SmallSpark Space Systems, Schultzy, and seven anonymous executive producers. Thank you all so much for your support for keeping this show around. Like I always say, those are real credits. If you want to put it into LinkedIn or something, I'll back you up that you produced this episode. But for now, that is all I've got for you. Thank you so much for

listening. Thanks for your support as always. I hope to see you in Florida, offnom.com slash events. Check it out there and I will talk to you soon.