EPISODE 143: NASA TRAINING SIMULATORS

Transcript

Andres Almeida (Host): Welcome to Small Steps, Giant Leaps, the podcast from NASA’s Academy of Program, Project & Engineering Leadership, or APPEL. In each episode, we explore the experiences and lessons learned of NASA’s technical workforce. I’m your host, Andres Almeida.

Behind every human mission to space, there’s a hidden layer of preparation that’s critical to NASA’s success. Today, we’re shining a spotlight on NASA’s Simulation and Graphics Branch at the Johnson Space Center in Houston, Texas. This team is responsible for creating high fidelity simulations and visual tools that helps train astronauts, design spacecraft and solve complex engineering challenges on the ground.

 Joining us today is Michael McFarlane, the branch chief of NASA simulations and graphics branch. Michael leads a team of experts who use cutting edge technology to create digital environments, tools and visualizations that support nearly every aspect of human spaceflight. We’ll discuss how simulations help astronauts prepare for space missions, the role of immersive technology like virtual and augmented reality, and how their work is shaping the future of exploration.

Hey, Michael, thanks for joining us today.

Michael McFarlane: Absolutely.

Host: Can you give us an overview of the Simulation and Graphics Branch and the areas you oversee?

McFarlane: Sure, Andres. So, the Simulation and Graphics Branch is all about providing simulations of space vehicles and space missions, really, for anything that NASA is looking to fly in the foreseeable future. Graphics is a big component of that. Obviously, we need to give the pilots as much positive feedback as we can, and that usually comes through their eyes. And so, we need to have good graphics. We need to have good simulations and good models all integrated together into a meaningful product.

And so, some of the areas that we integrate these products in – there’s one called the Systems Engineering Simulator, which is a complex in Building 16 here at Johnson Space Center, that incorporates dome visual systems and human-in-the-loop mockups. And we also have the Virtual Reality Lab over here in Building 16 as well. And we use virtual reality extensively for doing, really, mainly astronaut training in extravehicular activity operations.

Host: And I’ve been in that lab before, and I have to say, it does feel like you’re stepping out into space. It really does.

McFarlane: Obviously, that’s what we’re going for, so I love to hear your feedback. We love to get feedback from anybody that comes through, because that’s how we make a better facility.

Host: How does your team ensure that the simulations are high fidelity and also accurate enough for both engineering analysis and crew training?

McFarlane: Yeah, so I love the way that you phrase the question, especially the part about being accurate enough, because that is the key. When we do simulations and we’re working with a customer, there are certain models that need to be precisely accurate in order for them to be able to complete their assessment or study or whatever it is that they’re doing. Other models don’t need to be that precise. And other, models we can kind of fake out as the case may be. A big part of what we do is communicate with the customers to try to understand exactly what they’re trying to get out of the study so that we can do the proper level of verification and validation that would be required to get the results that are needed.

So typically, we are looking at orbital dynamics, robotic systems, and dynamics of feedback through those systems, mechanisms like berthing mechanisms and things like that, and we will work with the providers of those mechanisms to get whatever test data we can get our hands on. Ideally, flight data, if there’s flight data available. 

And we just dig in and look at plots, we look at data, we look at Excel spreadsheets, whatever we need to so that we can feel confident that we’re going to give answers that are accurate enough so that these groups can make flight decisions or whatever it is they need to, using the results from our sims. 

Host: And often, are these sims done strictly on laptops? 

McFarlane: So, that’s a good question. We do have the ability to run most of these. It’s basically the exact same sim that you would see in one of our dome facilities. You can run most of them on just a regular laptop. Sometimes a laptop might not have quite the power that you need to be able to run fully real time, but it’s, it’s usually close, and sometimes you don’t really care about running exactly real time.

But in fact, that’s one of the big changes that we’ve made over the last, you know, decade or two. It used to be that all of our sims ran on very high dollar, very, I don’t know, specialized equipment, I guess you would say, but now we can run all of that same stuff pretty much on any computer that you use to read your email.

Host: So how do your graphic simulations that your branch develops, how do they integrate with other areas of NASA operations?

McFarlane: We do a whole lot to try to coordinate with other groups that there are several groups that do graphics modeling, graphics rendering, across the agency. We try to keep in touch with a lot of them. So, we participate in or, in some cases, host working groups, getting people together just to really talk through, “Hey, what’s the latest and greatest technologies? What’s working well to achieve their goals and what’s working well to achieve our goals?”

Obviously, in recent years, virtual reality has been a big thing. There’s companies that are just releasing new versions of their headsets, you know, just very quickly, a lot of which have some really good capabilities that we care about, and a lot of them have capabilities that we don’t really care about. So, we try to talk through all that, try to share as much as we can.

There’s been a movement to get more and more stuff out on the open sim community, open source community, I should say, such that we can collaborate more easily with folks inside the agency and outside the agency.

And we’ve been very intentional about using commercially available products that are significantly cheaper now than they were 20 years ago. So, we historically have built our own image renderers, for instance, graphics renderers that, I mean, they do great. They achieve the results that the customer needs, but they tend to be very expensive.

Nowadays, with game engines and things like that, there is the capacity out there for us to be able to just use those game engines, just like any other game developer might, and incorporate them into our final products. And to be honest with you, those graphics rendering engines like Unreal and Unity and things like that, I mean, they have fantastic capability, a huge developer base, huge user base, and so we end up getting a whole lot of capability without really having to pay for it. And that’s a win for the government. It’s a win for the taxpayer.

Host: And you started during the shuttle era. I know this might be a big question, but what big changes have you seen in your career?

McFarlane: So, during the shuttle days, we basically had two programs. We had shuttle and station. Two programs was – it was pretty manageable. Nowadays, depending on how you count them, we have, you know, six to eight programs, which is, that’s a lot to, it’s a lot to balance. We found the only way we can do that effectively is by reusing a lot of our technologies.

And so, it used to be, we would have a shuttle sim that was almost entirely self-contained. Then when a new program sort of spun up, we would take bits and pieces of that shuttle sim, maybe, and use it as a foundation for a new vehicle sim. But that just doesn’t work anymore. We have to be much more agile, much more nimble in terms of how we create new content and new simulations.

And so, we build in reusability and modularity right from the very start in everything that we do. So, we have, for instance, an orbital dynamics package that we use over and over and over, because it really doesn’t matter what kind of vehicle you have, it’s the same orbital dynamics are going to apply. We also have tools like Trick. Trick comes out of our branch. It’s a simulation engine, and it helps us to really understand what can be reused in terms of scheduling jobs, in terms of data, recording, inputting data, modifying data on the fly, a lot of that type of stuff we used to have to do just kind of on a case-by-case basis. Now it’s all handled by some of these tools that we have invested in, and we’ve gotten other groups to buy into those tools as well, which, again, is an efficiency improvement, and it’s a win for the agency, interoperability is another one in terms of – Nowadays, we can interact almost seamlessly with simulations that don’t use Trick.

For instance, a lot of our partners, they have their own sim modeling engine, sim rendering, whatever sim operational engine. They don’t want to convert everything to Trick. And so, there’s international standards out there, like high-level architecture, there’s Space Federation Object Model or Space FOM or space farm. We can convince these partners to fully embrace those types of standards and technologies and, in doing so, we can almost just plug and play with their sim and our sim in a very straightforward and easy to integrate environment. Those types of things are a win for the agency.

Host: Can you talk about how feedback from astronauts influences simulation improvements?

McFarlane: Yeah, absolutely. So, whenever the astronauts come back from a mission, they did this during the shuttle days. They still do it today on space station, they have a crew debrief. That is one of the things that we absolutely attend, because almost invariably, they will talk about the things that they see on orbit and whether or not those things matched what they were expecting to see based on what they saw in the sim. And as you can imagine, they were not at all shy about letting us know if, if something doesn’t match, and that’s exactly what we want, we don’t want them to be polite and, you know, overly kind or anything like that.

If we messed up, we want to hear it, and so they’ll tell us, “Yeah, we were operating the arm, and the arm, you know, just seemed to, you know, be a moving a little slower than what we saw in the sim, or the feedback during this one particular operation was a little bit different than what we saw through the sim runs.” And we will tackle those one by one, and we’ll, we’ll dive in. You know, we got pretty good a relationship with a whole lot of the astronauts. They come through our facilities regularly. So, we’ll dive into it with them, try to understand it a little bit better. And usually, we can find the source of what the discrepancy might be. And now that said, we love it when the astronauts say it was just like in the sim. I mean, it’s a high five all around the room for us when we get that feedback. But that’s not always the feedback that we get.

Host: How do those sims differ from those for newer projects like Gateway that’s designed to orbit the Moon?

McFarlane: At its core, they don’t differ a whole lot. The station sims that we have have very old legacy. So, they, you know, we’ve modernized a whole lot since then, but a lot of that modernization, again, is in the form of modularity and reusability. So, you know, the modern sims like I say, when we are going to simulate a new vehicle — if somebody says, “Hey, there’s this new lunar lander concept that we want to consider,” well, guess what? We already have our orbital dynamics formulated. We already have the Trick sim engine formulated. We have generic vehicle control systems, generic life support systems, we have all sorts of these things modeled.

And so, what we will do nowadays is we’ll start from the very generic vehicle, get whatever information we can from the providers, and upgrade those models in kind of a modular setting. So, like I say, at its core, the physics haven’t changed, obviously.

We knew what the physics were of these operations way back when. The capacity of computing power has increased. There’s no doubt about it. So, we can understand some very complex dynamic operations a lot better than we used to. But for the most part. At its core, the sims of things like Gateway or lunar landers are very similar to the sims that we had of the International Space Station.

Host: I would love to know if you could share an example of something complex or a unique project that your team has worked on that really just required innovation.

McFarlane: We had a team, and it still exists. It’s called NASA Exploration In-Space Systems simulation, NExIS for short, is how that acronym plays out. But the NExIS team was really the one that grabbed the bull by the horns when it looked like we were going to have all these simultaneous programs. In fact, back about a decade ago or so, they had all these crazy ideas for potential missions we might fly. There was something that looked like an unmanned vehicle going out and sort of lassoing an asteroid and bringing it back to Earth. That was one mission that was looked at.

There was missions involving flying like a lander type vehicle to one of the Martian moons and bouncing around the moons and in order to do some science and stuff like that. So that team was the one that first, really, I would say, recognized the fact that we needed to be agile, and we needed to be reconfigurable and reusable. And so, they were the ones that sort of re-architected how we do simulations. It was very innovative. We had some of the smartest people in the world that started coming up with these various products that could be used just with a certain amount of reconfigurability.

And there was one called the Global Use Nodal Network Solver that we can use to model the flow of either fluid systems or electrical systems, just by using this, this basic engine, and plugging in components and various data to sort of characterize actual real-world systems on the vehicle. Oh my gosh, that has saved us just an incredible amount of time and energy in terms of developing new vehicles. And it’s also improved our confidence in the accuracy of those simulations that we have developed.

Other similar tools as well, there’s Valkyrie, which we have set up as being a generic GNC (guidance, navigation, and control) modeling sim. There’s Hydra, which is a generic robotic arm simulation. We have input device framework that allows the astronauts or the study groups to tie in almost any input device, like a hand control or steering wheel, whatever case it might be, in a very simple way. We used to struggle with that. Anytime we got a new piece of hardware, we would just start from scratch. Now, because of the investment that that NExIS team did, we can turn things around very quickly.

Host: How about with all the advancements in technology and augmented reality, virtual reality and AI, (artificial intelligence), how do you see simulation and graphics evolving at NASA?

McFarlane: I mean, that’s something that we’re constantly asking ourselves. So we are tied in very tightly with these industry partners, trying to stay ahead of the curve in terms of the technologies that they produce, and really kind of feeding into them to say, hey, here’s some things that we think would be a win, certainly for NASA, but potentially for other customers as well.

You know it’s, boy, I don’t want to get too deep into the weeds, but one of the things that we found out is when you fly these VR headsets in space – a VR headset actually needs a gravity vector to understand how you rotate your head around and how that feeds back into the system. Well, guess what you don’t have when you’re flying around in orbit is you don’t have a gravity vector. So we had to work with those suppliers to come up with an alternate plan. There are other technologies we’re working on with those guys as well.

You did mention AI. AI is obviously, that’s the technology that’s kind of the hot thing these days, and we have already had some individuals that have taken some of the Trick software, and Trick is open source, so they could do this. It wasn’t. There’s nothing proprietary about Trick at all. So, they could feed it into one of the commercial AI generative AI packages. And they just ask it, “Hey, build a sim of this, that or the other.” And I mean, it came up with something that was usable. And so, I mean to me, that’s going to be the future is that you’re just going to tell your computer, “Hey, computer, I need a sim of the Earth and the Moon and this vehicle that goes back and forth, and it’s going to have to dock to this other vehicle,” and you’re going to push a button and say, “Go,” and it’s going to get you 80% of the way there. You’re still going to have to do some fine tuning and stuff like that, but that is going to be really exciting thing for everybody involved, to be able to improve how the development of simulations happens.

You know, obviously one of the big steps there is going to be having the ability to have it ingest sensitive data, which we which we can’t do at this stage. So, they’ll have to figure out some way of firewalling off sensitive data so that we can let it ingest that we’re not there yet.

Host: Michael, did you have mentors in your career? And what advice would you give to early career engineers who aspire to go higher at NASA?

McFarlane: Yeah, I had some at the time. They called us part of the Cooperative Education Program. So, they called us co-ops. Nowadays, it’s basically the same program. It’s called the Pathways intern program. My gosh, that was one of the one of the best things that ever happened to me, was becoming a part of that program. And I almost just stumbled into it too. I got a little bit lucky.

We had some guys come to talk to us in one of our classrooms at University of Michigan, and, you know, I thought, “Holy cow, is this even real? I, you know, I could be an undergrad and work at NASA for NASA?” So I, you know, I got in my name, and somehow I got picked and, I mean, the rest is history. It was just a fantastic program.

During my time as a co-op, I had a couple of great mentors. One was Andre Sylvester, and the other was Mike Read. And you know, the way things work out, they both ended up being branch chiefs for me, so I had ended up with long-term career interactions with them. And I mean, they were great, you know, they helped me to see my potential. They helped me to understand how NASA works, how you, how you do business at NASA, and how you can be successful.

And, you know, you kind of asked advice that I might have for young and early career folks. So one is, get yourself or find a good mentor. If you’re not in a program, and there’s good programs out there for mentorships, you can do formal ones. You can do informal ones. Chances are, if you talk to a branch chief or division chief and say, “Hey, you know, can I use you as a mentor? Can we, you know, maybe meet once a month and talk for a half hour.” I think most division chiefs and branch chiefs are going to say, “Absolutely!” So yeah, I would say, do that.

But more specifically, too, I would say, understand how business works in the federal government. Federal government is just very different than the way the rest of the world works in terms of the way funding flows, in terms of the way responsibility flows, and it is very hard to decode. And a lot of, especially a lot of young engineers, will say, “I don’t want to deal with it. It just seems too confusing. It seems irrational. So I just, I’m going to wash my hands of it,” but I would say don’t do that. It’s really not as difficult as it seems, to understand how business works within the agency. And once you understand how it works, you can start the funding flowing, and you can start turning your ideas into reality, which, I mean, that’s the best thing you can do as an engineer.

Host: What would you consider to be your giant leap?

McFarlane: I knew that question was coming, and it’s a difficult one, but, you know, I’m not sure if I had a giant leap anytime in my career, but I’d say that I had a lot of small leaps that all were kind of fairly well-aligned.

Just growing up as a kid, I was a big Star Trek fan, you know, The Original Series with Captain Kirk and Mr. Spock. And one of the things I loved about watching those shows is I thought, “This is stuff that’s achievable.” I mean, you know, maybe not in our lifetime, but I thought, this is just engineering. These are all problems that at some point in the future of this country or this world, we’re going to solve all these problems, and I wanted to be a part of that, you know.

So it’s goofy to think a TV show could really change the course of somebody’s life. And I had great parents who saw that I was that interested in the show and really encouraged me to seek out math and science and physics and stuff like that. And eventually, an engineering degree from University of Michigan that opened up all these doors for me. Since that time, it was really just a matter of taking leaps when opportunities presented themselves,

I didn’t have maybe a really clear career path mapped out for myself. That’s one of the questions supervisors always ask all of their employees: What’s your five-year plan? What’s your 10-year plan? I didn’t have great answers for those, but what I knew was that opportunities would present themselves, and I wouldn’t jump at every opportunity, but I would, you know, I would assess every opportunity, and every once in a while, good ones would come up, and I would, I would go after them hard. And you know, when the opportunities do present themselves to you, don’t, you know, don’t go halfway in on them, go, go all in on them.

And that’s what I did throughout my career, just, you know, leadership opportunities, management positions, whatever the case may be, if they were right, I worked hard to make sure that I had a good shot at securing those responsibilities.

Host: So, go for it, is what you’re saying.

McFarlane: You got one career. You don’t want to have just an ordinary career, you know? You want to, you want to make a difference by the time you retire.

Host: Yeah. Great advice, Michael, thanks so much for your time.

McFarlane: Absolutely, Andres, I really appreciate you talking with me today.  

Host: That’s it for this episode of Small Steps, Giant Leaps. For more on Michael and the topic we discussed, visit our resource and don’t forget to check out our other podcasts like Houston, We Have a Podcast and Curious Universe. Thanks for listening.