Naeem Altaf | IBM Space Tech CTO, D.E. | In the Open with Luke and Joe

Luke Schantz: What you're about to hear is a new podcast and livestream show entitled In the Open with Luke and Joe. In this series, my co- host Joe Sepi and I bring you conversations with community and technical leaders from the world of open source and enterprise tech. We do this live twice a month on Fridays at 12:00 noon Eastern Time. You can catch us on a variety of streaming platforms or hear as a replay on your favorite podcast app. To find out all the details, go to ibm. biz/ intheopen. There you will find our show schedule, an embedded player of the live streaming video, as well as embeds of past episodes. Or you can link directly to the podcast page with ibm. biz/ intheopenpodcast. Thanks so much. I hope you enjoy our new series, In the Open with Luke and Joe. Welcome to In the Open with Luke and Joe. I'm your host, Luke Schantz, and here's my co- host, Joe Sepi, and a big welcome to our guest, SpaceTech CTO Naeem. Before we get to our show, don't forget to like and subscribe. Thank you for joining us today. We have an exciting show for you. We're going to be talking about SpaceTech and edge computing, but before we bring in our guest, Naeem, let's bring in our co- host, Joe Sepi. Hey, Joe.

Joe Sepi: Hey, how are you?

Luke Schantz: I'm good. Welcome back. Glad to have you back on In the Open.

Joe Sepi: Yeah, it's great to be here. Second show, rocking and rolling.

Luke Schantz: So... Oh, go ahead.

Joe Sepi: The weather is great here and I know you know that, too, because you're in Connecticut as well, but it's like it's 58 degrees. It's fantastic.

Luke Schantz: It is. I feel like winter may give us another little taste I think, but I am feeling that the hope springs eternal. This is very... It's a beautiful time of the year. I think there's

Joe Sepi: For sure.

Luke Schantz: ...even been some space weather lately. I saw something in the news about a space hurricane. We're going to have to ask Naeem about that.

Joe Sepi: Yeah, fascinating.

Luke Schantz: Yeah. This is going to be such an interesting show because edge computing is a big topic that I'm interested in, and it seems to me like edge computing in space is the ultimate edge, so the edge of our known world.

Joe Sepi: Yeah, and one of the things that kind of... I don't know if pride is the right word, but having joined IBM a few years ago, the work that we've done with NASA with the moon landing, I love the SpaceTech stuff. I'm excited to talk to Naeem.

Luke Schantz: Before we bring him in, he gave me a short video that we could play that's going to set up one of our first topics we're going to be talking about, so let's get to that video and without further ado.

Speaker 3: From the earliest days of Apollo when NASA landed the first humans on the moon, IBM has had the distinct honor of playing a key role in NASA's space exploration efforts. Fast forward 50 years to today when the International Space Station is helping to lay the groundwork for living and working in space, a baseline for future NASA missions. Traveling at over 17, 000 miles per hour and orbiting the earth every 90 minutes, the ISS offers a one- of- a- kind microgravity environment where crew members perform research that cannot be done anywhere else. Critical research like DNA sequencing on the ISS provides foundational knowledge that will be essential as NASA seeks to venture further into space than ever before. However, analyzing this research often requires data to be downlinked to Earth and processed by personnel on the ground, a procedure that can take several weeks and delay results. That's where IBM comes in. In partnership with NASA, ISS National Lab, HPE, and Red Hat, IBM created the edge computing in space solution, eliminating the need to move massive libraries of DNA sequencing data by presenting containerized analytical code locally right on the ISS where the data originates. This solution has the potential to cut analysis time by nearly 50%, opening the door for many new mission possibilities. In addition, NASA researchers will use this platform to more rapidly develop, test, and push code to the ISS in a fraction of the time by leveraging Red Hat code- ready containers and connecting to IBM Cloud running OpenShift on the ground. This groundbreaking partnership will not only expedite NASA's ISS research, but will help to lay the foundation for future exploration opportunities on ISS and beyond. We can't wait to support what comes next.

Luke Schantz: Hello, welcome to In the Open, Naeem. Thank you for joining us.

Naeem Altaf: Thank you very much for having me.

Luke Schantz: Before we dig into so many exciting topics today, let's just help our audience out with a brief self- introduction so they get to know who you are and where you're coming from.

Naeem Altaf: Sure, so yes, my name is Naeem Altaf, IBM's distinguished engineer and CTO for SpaceTech. I run an innovation lab which is based here in Austin, Texas.

Luke Schantz: Excellent, and maybe we should lay a little foundation here, too, for our audience if maybe we need to let people know what edge is and why. Obviously, the thing that comes to mind here is bandwidth issues. You can't have all the data and space and bring it back down, so I think this is actually if people weren't familiar with Edge, this is maybe one of the best use cases to help you understand why edge.

Naeem Altaf: Sure. Yeah, so let's start from the terrestrial on the ground stuff. There's a huge buzz in the industry about the 5G networks, so basically at a very high level what edge means is that or what edge computing means wherever the data is being produced, you do computation right there. Let's say if you have a sensor far away in the ocean which is looking for temperature, you don't want to bring all that data onto wherever you're on- prem on the cloud and do processing. You want to do processing right there, so that's what basically the edge computing means. Another good example is the smart cars and the smart home of the future and the autonomous cars. Imagine if the car and autonomous cars of the future, every time they have to make a decision, they're asking the cloud or coming back to on- prem data center and saying, " What should I do now?" Can you imagine what can go wrong then? It has to make a decision in milliseconds right there, so that's like a moving edge computing platform right there where it's making decisions, learning from its surrounding from all the sensors and making a decision right there.

Joe Sepi: Essentially, it's really like decentralizing where the computational stuff is happening and which is really important with all the IoT devices that are proliferating around the world.

Naeem Altaf: Exactly. Big thanks, too, because the compute is very dense now. The microcontrollers we have like Raspberry Pi and inaudible the Jetsons, Nano Jetsons from NVIDIA, they are so powerful. They even have GPUs on them, so you can do inferencing, you can do modeling, so whatever in the remote areas you can do that. In our case, we wanted to extend that to space because, like you mentioned, the latency issue, the bandwidth issue, if we have challenges on the ground, there are much bigger challenges when we are in the orbit around 5, 600 kilometers above. There's a lot of stuff which happens there. We'll talk about that, and the idea was, can we do competition right there and get the actual level insight and sell that piece of info nugget?

Luke Schantz: That makes so much sense, so walk us through a little bit more detail than what was in the video about-

Naeem Altaf: Sure.

Luke Schantz: ...what's being done now and maybe how it's going to be used in the future.

Naeem Altaf: Yes, so last year we got together with the International Space Station, National Labs LASSO and HP and we were looking for the projects where we can extend the concept of edge computing in space. They mentioned this project DNA sequencing, which runs in the space station. As a space station, in simple words it's a big laboratory which is orbiting the Earth. There's so many experiments happen for microgravity for the future exploration as we are heading towards the next few years towards Moon and planning towards Mars. This is like a lab in the orbit so you can do just different tests. One of these use cases of DNA sequencing was where they take sample from the surface or they're looking for microbes in the environment and they can do a DNA sequencing right there in the orbit. I think three years ago they sent this midiron device up there, which is a sort of a handheld device which can plug into as a USB device into a computer. The way it worked for that was last year that the PI, which is the principal investigator on the ground, they will put out a question and they'll say, " We want astronauts to take a sample." Based on whatever the schedule of astronaut, they'll take a sample and it can create up to half a terabyte of data for its one run. Then, all of the data needs to come down, and then it needs to be processed and all the logistics and everything. A significant time has passed since I put that request in and when I my results back. What I'm looking for is basically a resulting file like a PDF file which tells me that resulting data. We said, "Okay, you know what? We can do all of this processing right there," because on February 20th this year, the SBC- 2 computers from HP, they were going up in as spaceborne computers and we were in partnership with them. They know that we have the compute right there which has GPU and CPUs, so all we need to do is take this open source code and it has like multiple steps. It's very complex. It has the base scaling, it has the multiplexing alignment analysis, so you broke it down into, okay, for the base scaling, we need a much more powerful processing, this GPU. We used GPU for that, and then when the results comes out, we'll take a second system because there are two systems for the CPU computation. We took the open source code, we containerized it, and then we use our OpenShift code- ready container platform. We package that whole thing. Of course, there's a lot of testing, security scans you have to go through on the ground because we have the exact replica of those two systems on the ground at NASA's facility, so we can test it before you can push it up. We did all the testing, the stuff. The code was put on, I think, around October on the flight system, and now they're up there, so we are hoping by end of this month or first week of April we will be able to bring systems live up and start running through the process. Just to close, the whole process will take around six to eight hours. Let's say if PI put that request in and next week it's scheduled for the astronauts to take a sample, literally after that we'll take probably a day or two and we'll have the results.

Joe Sepi: That's as compared to what before-

Naeem Altaf: Probably like six to eight weeks is... yes.

Luke Schantz: That is huge. Something Joe and I were talking about, Naeem, when we were prepping for... This is also exciting and I must say I have serious jealousy when I hear you talk about, " Oh yeah, I was working with the space station and they asked me if I could do this," and I feel... I like what I do, but that is definitely cooler than I think what I do. Let me ask you this, is there any way that the average developer or student could get involved and somehow be doing something with space computing?

Naeem Altaf: Definitely, so let's say, for example, if you go onto the NASA websites, they put out these different projects all the time for the community to get involved. They're also big in open source. IBM is big, huge on open source. We open- sourced two projects last year which was recognized as a top five open source projects for the game changing of future. One was about the space situational awareness and one was about the CubeSat frameworks for the autonomous CubeSats or drones. It's open to public. Anybody should be able to hit them and inaudible work together.

Joe Sepi: Yeah, that's really cool. I'm super excited about this sort of work. I get these alerts when the ISS is flying over my sky and can go out there and take my son out and that's... They're out there in space flying over, so it's really cool and it's great that folks can get involved with that work. I'm really excited.

Naeem Altaf: Definitely, and if you folks get a chance, and we'll talk later about a little bit about inaudible. It's like a train of 60 sats. It's like a reindeer passing by. You cannot miss it, so go on the site. There are any sites which tells you when these lights will do a pass over your location and you can see these things. It's really cool to see that there.

Luke Schantz: Yeah, and that's a good segue, too. I mean, there's lots of stuff orbiting the Earth and I was reading a little bit about... The link is down there too, the SpaceTech SSA. Talk to me about that because it's amazing how much stuff is out there from small to really large and traveling at great speeds and it seems like it'd be a little bit of chaos.

Naeem Altaf: Yes, so by the way, the speed in the lower Earth orbit, which is around 300 to 1, 000 miles, that's classified as a lower Earth orbit. They're traveling at 17, 500 miles per hour, these objects. It is extremely fast, and even if you have a small debris, 10 centimeter, one centimeter, it can be more substantial damage if it collides with an object. This is a very hot topic in the space industry, and you will see the names like space situational awareness, space debris, space traffic management, but still we are trying to have a grip on what's out there in the orbit. Now, think about you have seen these sites like Flightradar and stuff like that where you're tracking your flights going from point A to point B. Imagine if your control tower is saying that a flight coming from Austin to Connecticut and it shows you four paths. Right now, it can be a disaster on the runway. You don't know which one to trust, which is the right path, and you have thousands of flights every day. This is a problem where we need to track what exactly is there and, what is the path of these different objects? The way today it's done is with the radars and it's done through different Department of Defense and there are some private companies doing that. Going back to your point, what's out there? What's this like? Today, before Starlink came two years and they started launching these satellites, we had launched 9, 000 satellites which the public knows of, and out of those 5, 000 are active. The rest is debris and stuff like that. Now, just today there was another launch from SpaceX to put these Starlink satellites this morning, and their count, I think, has gone to 1200-plus now. They plan to put 14, 000 satellites in the next two, three years. Then, there is Amazon Kuiper. They want to put 3200 there. Then there is OneWeb. They want to put another 2, 000. Then, there is an unknown company, it's still being talked about from China, they want to put 13, 000 more. You're talking about around 30 to 40, 000 sats in the next three to five years. Whereas, to date you only had between 5 to 10, 000. Can you imagine what's going to happen in that congested space? That's the commercial reason why they want buy broadband all this stuff. That's why the Conjunction Search, so we want to know when the objects are coming close to each other. Just last year, space station had to maneuver three times. Maneuvering in the orbit is like the last resort. You don't want to burn your fuel. It's very limited and expensive, and one of the part from our previous rocket, it was coming within a few kilometers range. Actually, the astronauts had to go into the evacuation mode I think into the Soyuz capsule to exit in case it was that serious. I think this is where the open source community, the developers, can really take you to the next level, and I would even go as far as saying this is an optimization problem. This is where quantum... These are the magnitude of problems that quantum can help, even though it's the early stages of quantum computing, but start thinking about it. What can be done? How can we use this next generation of compute to solve these problems? Another thing is it's causing also lots of challenges for these astronomers on the ground. You have seen these billion- dollar telescopes in South America and Chile and Europe which are doing these deep space observations. Now, imagine if we have this web of stuff in their path. You are obstructing with the data light coming back in, but so we have to think about it. How can we optimize all these things and everybody's happy?

Luke Schantz: Yeah, and that doesn't seem... Our producer Scott is blasting messages at us, so I think it is apt, like massive air traffic controller problem. This sort of stuff isn't managed by humans, right? I would have to think that this is like AI/ ML sort of work, and you even mentioned quantum. How do you manage, especially with the proliferation of stuff in the future? How do you really manage all of that?

Naeem Altaf: It's hard to know what's in the orbit just for the commercial companies, even though there are few which are emerging. Mostly this domain is owned by the DOD and the forces because they have these radars and all this stuff. What needs to happen, first of all, this is a policy thing where the governments needs to come together and say this is a serious threat for humanity. We have to come together and work just like climate change kind of stuff. We need to work together. We need to have policies and technology is there. Technology is not a problem. For example, we can use Blockchains. If let's say all the agencies, they start sharing data, we can use Blockchain, so we can have the transparency, providence, trust built from there so everybody knows who is touching what in the datasphere.. Then, we can have AI/ ML to figure it out to detect these orbits and predict better movement of these objects. This year, I think this year, yeah, iSSA is sending out a spacecraft to clean up, to remove that debris. That's a classic traveling salesperson problem. Again, looking at supercomputing, potentially quantum in future, it can be done, but the governments have to come together. What I fear is like any other thing, they probably wait for something happen and then say, " Oh, now we have to get together." I hope before something bad happens, we come together and pay attention. This is a serious problem.

Joe Sepi: Yeah, it's the classic kind of thing. Code is easy, people are hard. It's not just the code, it's the government entities and policy and getting everybody on board and working together. I imagine that's probably not easy, and then I don't know if this is a question from the chat or from our producer here, but it's a good question. Are there multiple orbit layers like high Earth versus low Earth-

Luke Schantz: Varying-

Joe Sepi: ...inaudible opportunities and challenges within that?

Naeem Altaf: There are three layers on the top. One is a lower Earth orbit, one is a middle Earth orbit, and one is a synchronous orbit. In the lower Earth orbit, that ranges between 100 to 1, 000 miles, and that's where most of your Earth observing satellites, space station is about 254 miles above. Majority of the stuff is in lower Earth orbit, and since it's moving so fast, that's why you can see them orbiting almost 12 to 13 times a day, these satellites. For Earth observation, these modern satellites which should go up, they can scan the whole globe right now in two days, in future, in one day. Your data is only for 24 hours, so imagine that, and the resolution is 50 centimeter. They can literally watch what's in your pocket kind of stuff. The second is the middle Earth orbit. That's where your GPS satellites are there and military stuff is there, and then you have the geo, which is the above 22, 000 miles above. This is where if you have the dish at home like Dish Network or dish, because it's synchronous with the Earth's rotation. It's not like low Earth orbit it's going zoom, zoom, zoom. You will lose connection every 10 minutes because the pass is only eight to 10 minutes. In geo, you are in sync. You have dishes, so you always have a constant connection. These are the three orbits, yes.

Luke Schantz: Fascinating, so let me ask you about the other project we're linking below here, the CubeSat project. Where does that fit into the orbits? Where does that fit into the hierarchy of CubeSats? I'm imagining you've got these space telescopes and communication satellites at the top, and then we've got this whole continuum down now to the CubeSat.

Naeem Altaf: The idea behind that open source CubeSat project was let's get a distributed computing, and if you have a swarm of these CubeSats. What's happening in the industry, and I think from our developer point of view in the cloud- native world, they can probably related to this. Before the cloud- native containers cube, we had these monolithic products. Takes long time to develop and have very less frequent releases and all of that stuff. It's a big, big thing. Think of that, those big satellites, which we build for 2 to $ 300 million, takes three to five years. Huge cost and they send it up and they stay there for 10, 15 years and that's the end of life. With the concept of CubeSat, it's very similar to microservices. You have these very small, lightweight, cheap... I mean, for all cube sensors and everything is half million compared to 2 to 300 million. Students build these CubeSats for less than$ 50,000. Let's say even if you take a pro staff, half a million dollar, you can build these things within six months. Within one year, you can launch that. If you have these swarm of these CubeSats and because the technology is so advanced, the cameras we have and the compute power, you have a Raspberry Pi with inaudible memory with quad- core processors and everything and GPUs as well with other microprocessors. It's very powerful. What's happening in the industrial trend we are watching, the companies are going with these small sat category like CubeSats, even nanosats. What we thought, "Okay, if we have a swarm of these, let's say 13 of these, we launch them." There is a hurricane which is coming in the Atlantic, so basically it's fully automated, fully autonomous. A signal goes from the bottom or the machine learning model detects that something being formed in the ocean. It asks all of its peers. Basically it's based on messaging system, open source inaudible. It asks, " Okay, who has the bandwidth in terms of resources to go and look at this problem," Out of it, we are available. It forms a virtual cluster and it'll start working towards this problem. Then, you can get the payload from the ground and say then, " Here's a new model. Start doing inferencing on this kind of a problem." Basically, distributed computing at the edge. You're forming clusters to tackle that problem. The same concept can be applied on the drones. We can have cluster of these drones in the ground. Actually, that brings you another point is a very interesting thing happening in the industry. The next generation of communication between the ground and satellites is the optical. It's a laser communication much faster, high bandwidth. Today, we do mostly radio signals, but the challenge over there is clouds, and most of the time the Earth is covered with clouds, the areas which of interest. How do you handle that stuff? There are companies who are saying, " You know what? We can build these drones which can fly above the clouds so they can be like a middleware." They will intercept the signal from the satellite and they will do processing, and then they will fly to where they have a line of sight to the ground station clear and then do a downlink. A lot of innovation is happening, but I think these light optical connections, it's going to move even more compute up there because you will have so much fast connections. See, the compute is not a challenge. You saw with Mars Rover, we are using the IBM's chip power PC chip from 10, 15 years ago, 10, 15. The network bandwidth and speed is the challenges. If we solve that problem, it'll be really cool.

Luke Schantz: That is amazing to think about. It really is like some kind of science fiction scenario you just laid out about beaming data to drones that are relaying it.

Naeem Altaf: Yes.

Luke Schantz: It reminds me like I saw in Eastern Europe where they didn't have a lot of infrastructure. This is 20 years ago, they were building open source line- of- site IR network devices. They would just use a piece of PVC and literally like the IR generator from a remote control with a lens and they could create these one- kilometer like 10- megabit connections, but again, like you're saying, once it gets foggy or is raining the network's down.

Naeem Altaf: Yes.

Joe Sepi: Yeah, people talk about clouds, moving everything to the cloud, but now clouds are in the way.

Naeem Altaf: Yeah, you move that cloud up into the orbit, right?

Joe Sepi: Yeah.

Naeem Altaf: So inaudible-

Joe Sepi: Yeah, yeah-

Naeem Altaf: ...inaudible-

Joe Sepi: ...when you were talking about drones, I was like, " Is it going to blow the cloud away?"

Naeem Altaf: That that can be interesting, too. Have a much bigger drone with a fan and say, " Just move aside."

Joe Sepi: Yeah, at least enough to get the optical connections. It's fascinating.

Naeem Altaf: That also brings up... Again, these are new problems. It's very challenging, interesting stuff. Optimization path for the communication from the orbit, the ground station. Again, opportunity for machine learning, AI potential, quantum in the future because you cannot have ground stations like the cell towers. It will not look pretty if we had 50, 70,000 grounds. If you have only limited one, how do you do route efficiently? How do you route it down and how do you route between them? This is an optimization problem.

Joe Sepi: Yeah, and like you said, the traveling salesman. I imagine we talk about quantum emerging over the coming years and I wonder how much more that will come into play with these scenarios that you're describing.

Naeem Altaf: Yes.

Joe Sepi: It's really interesting.

Luke Schantz: Naeem, you were on my podcast last year, and when we were talking-

Naeem Altaf: Yes.

Luke Schantz: ...then, you had mentioned there were all of these Mars missions coming up within the next six months. Now, I think all of them have played out, so could you give us just a quick little industry snapshot of what's happening with Mars? Help give us some context around why it's important.

Naeem Altaf: Sure. Yeah, so last year around summer in, I think, July or August timeframe, these three missions flew. The first one was the UAE Mars Hope, and that was an orbiter. The second was detained, ran from China. That was orbiter plus the rover and the lander. Then, the third one was The Perseverance, which was the lander and the rover. The first one, of course, it takes six months to travel, and in February the first one was the Hope Probe. The second one was the Chinese probe and rover, and the third one was the U. S. rover. The U. S. probe Hope is orbiting and it's primary goal was to look at the weather patterns on the Mars and because that will help in the future if you want to inhabitability and it's sending people to the Mars in next decade or two decades. It's going to learn about the atmosphere of Mars. The Chinese, similar thing, it's looking for ancient life. They will deploy their robot in May and that is the most hardest as you obviously saw the seven minutes of terror, as they say, the minutes coming down because the lag is 11 minutes of communication between Mars and Earth. If I told you something, " Go and execute this," for the next 11 minutes, I have no idea what you did, and within those 11 minutes and seven minutes, it has already landed. China will attempt that in May, and so they're also orbiting Mars and looking for the atmosphere, and then they will go to look for ancient life. This is also very interesting because they have a helicopter inaudible also attached underneath that rover, so they are doing multiple things. They want to also go and look for any signs of life, and the very interesting thing, they are going to collect these samples and then hand over to this helicopter, and the helicopter will go and then there will be a future rocket come. Will pick up the sample and bring it back. That's to be in future, but that's what the plan is about, so yeah, very fascinating. I think it made some movements yesterday or two days ago, and it can hear all sorts of inaudible we have much more interesting the computers on the top. The chip was as was the power PC 750 single port. If you just want to do a comparison, so that chip is 233 megahertz and the clock speed you have on your iPhone is 3. 2 gigahertz. It's comparing 6 billion transistors to almost 16 billion transistors. You can imagine how much difference is, but the name of the game is liability because it has to be radiation- hardened. The environment is very harsh. Anyways, this is glad to have that chip over there and running, but in near future, we'll start seeing some very interesting things coming back. You already see the pictures coming from The Perseverance, right?

Joe Sepi: Yep.

Naeem Altaf: Very fascinating.

Luke Schantz: Amazing, and there's talk of within decades or decade even sending people to Mars. Does this seem realistic to you? I know that's a out of the blue question there, but...

Naeem Altaf: The biggest challenge first is the landing. The landing is the hardest part on the Mars because of its atmosphere. As you saw, this time they use parachute to land the rover. Now, if you are talking about Starship, it's a gigantic ship, almost like 10 stories tall. I think in next 10 years if they can successfully multiple times launch and as the window is so long it's six- month window to get there, so I think maybe after 10 years because this robot is also going. The atmosphere is full of carbon dioxide, so you have to convert that CO2 into the oxygen. I think 20, 30 years, yes. That's far.

Luke Schantz: So interesting.

Joe Sepi: Yeah, it's fascinating. I want to bring in a question from our producer Scott, and this was something I was thinking about earlier, too. Scott asked, " Are there any concerns about a critical mass of orbiting devices and crafts that somehow impacts like weather systems, global temperatures, reflected sunlight?" Which is something I had wondered about, too. " How can we help monitor that and-

Naeem Altaf: You mean for the inaudible the orbit, correct?

Joe Sepi: Yeah, the orbit inaudible.

Naeem Altaf: Yes, so it's a big concern, like the one concern we heard from the astronomers and the scientists on the ground was that the light which is being produced by the Starlink satellites because of so many of things is reflection and causing disturbance in the data, but the solar weather, so you guys were mentioning earlier that hurricane which we just saw this phenomenon. First time we have witnessed that. Basically, you probably are aware of aurora when the solar winds or the solar activity happen from the sun and it hits the magnetic field on the North and South Pole. When they're both combined, this beautiful thing happens in which we can the Northern Rains or Northern Lights they call them. Yeah, it's a huge concern. Actually, it's a very good question you asked. Our next open source and research project, which we just kicked off last week, is going to be space weather because space weather, there's a huge impact, the communication. You probably have heard about that in certain regions the world, the solar communications were not available for this amount of time. This is because your sun becomes so active and it sends those flares. We can get like to to three days in advance. NASA has the website where you can get all this data, and based on that, it will damage the electronics over the air, but hits directly on the ground as well where it can damage the grid. This is serious stuff and we're going to start working on the space weather now, and again, I'll put it to the community who wants to help, but this is another interesting use case and it definitely has an impact.

Luke Schantz: I remember the scenario like this that happened in, what, the late 1800s they called the Carrington Event where there was some sort of solar weather that took down the telegraphy system, and maybe we haven't had an event-

Naeem Altaf: Yes.

Luke Schantz: ... ofthat equivalence, but we actually experienced solar weather affecting telegraphs.

Naeem Altaf: Yeah, and now we have the technology, we have the probe. We have... There is a solar probe which is so... Just it's mind- boggling that in this probe, it's so close. It's watching the sunlight, so we have probes between the sun and we have, I think, at least two or three layers of these probes which are watching the solar activities. That can help us to predict the events happening on the Earth and how you can maneuver. Again, like I said, in next three, five years, if the skies are going to be covered with all these 50,000 servers, the first thing is the servers are going to get affected. Everybody needs to be smarter on how to use these resources because the space is for everyone. It's not for one person, it's for all of us, so we have to be very smart how we use these resources.

Luke Schantz: You had mentioned, too, the regulation around this is... I guess my question here is this seems like a scenario where it's almost like the age of the ocean from a bygone era where you have the law of the sea and, like you're saying, it's based on precedence. It seems like obviously we've been in space for a while, less than a hundred years, but a lot of this is like you're saying, it's uncharted territory literally and figuratively that we've got to work out with these different countries for sort of a collective good.

Naeem Altaf: Yes, for example, if you have a half- a- billion- dollar satellite in low Earth orbit and I have my CubeSat and they're coming in each other's way, who should move? Common sense will tell you who should move, but so a similar incident happened without naming the two companies last year. One had a big satellite and one had the pizza box and they said, " Move because in the next three days there's a possibility of collision." The messaging which is happening is through... I sent you an email and the pizza box company said, " Oh, we never saw your email." Of course, the half- a- billion- dollar guys had to move their stuff. They don't want to get it destroyed, so you can see this. Yeah, there needs to be policies, lots more kind of for a real- time system. We have fixed this problem with the social media with the amount of traffic which flows every day. This is a very small problem compared to that. I think it can be right, it's just a matter of people coming together.

Joe Sepi: When you say pizza box company, I'm just imaging, you know, " Thank you for ordering" box flying through the sky. What inaudible-

Naeem Altaf: It's just like very small, but there was a company I think last year or the year before, they said they want to advertise from the orbit. They want to have a big display. Come on, let's not go that far because there's so much great work happening on the ground for deep space observation. We don't want no ads showing up at the middle of the night from the lower Earth orbit, but there's lots of crazy ideas happening right now. Thank you.

Joe Sepi: I'm reminded when I lived in Boise a long time ago, there was a local pizza place that actually delivered a pizza to Alaska and it was a big story, but I'm curious though. As there's more of like kind of consumer sort of stuff, how do you manage that? Maybe that's off- topic, but it seems like the proliferation would just be astronomical, pun intended inaudible-

Naeem Altaf: inaudible the regulatory authorities, they are the only ones. Otherwise, it's the Wild West. Just go and claim and the next thing is Moon. Okay, whoever goes there first, put their flag. It's theirs, the whole Moon is theirs, so there has to be a genesis coming together and the regulatory authorities and have a plan for it. We are not opposing the technology. It's great for everyone, but let's have a plan. Let's not pollute it. Yeah.

Joe Sepi: Yeah. I was trying to explain to my son last week. He was talking about selling property on the Moon and I'm like, " You can't just sell it." Anyway, that's just reminded of that.

Luke Schantz: That's interesting. He was going to be a... He's already... He's enterprising. He wants to be a real estate agent on the Moon, so-

Naeem Altaf: Yeah, add one more thing on that note. The astromining, I think that's a trillion or... I don't know what's higher than a trillion. I know it was a very... some interesting word which has many zeroes I can't count. That's the industry to go in because go and get that rock from that flying asteroid. We already saw attempts last year, the Japanese Agency and the NASA, they both took a sample and brought back. That's another huge area of exploration.

Joe Sepi: You combine that with NFTs and forget it.

Luke Schantz: What?

Joe Sepi: Yeah.

Luke Schantz: I'm going to short asteroid mining.

Joe Sepi: Weren't we talking about some sort of astromining thing a long time ago, Luke? What was-

Luke Schantz: Yeah, we were at an event I think on Roosevelt Island at that tech campus and I had just read about it and I came to you and I'm like." Joe, this is a trillion- dollar idea. We're going to do space mining and I'm going to set it up for IBM. A trillion- dollar idea." I think people were looking at me like... they were like, " What is he talking about? Should I be listening to this?" I'm still into it. I want in.

Joe Sepi: Yeah. Yeah. That's funny.

Luke Schantz: One question that came to mind about the idea of edge on satellites we're talking about now, too, but I could envision now that we're going to be going more to the Moon and Mars, are we going to see a data center on Mars where, let's say, IBM puts a data center on Mars and we extend that model? Is that what's going to happen?

Naeem Altaf: So-

Luke Schantz: Okay.

Naeem Altaf: ... innext four years, yes, we are at the Artemis Program going back to the Moon. The Nokia, they put out a contract to put 4G network there and a NASA team, they want to put a data center. There are already discussions about lunar gateway, so yes, it is happening, so putting data centers because this time their mission is to go and stay and to set up a base there for future launches from there, especially if they can find water over there. They can use that with other fuel and stuff to go to the next one. That will be huge, so yes, data centers are coming on the Moon if this Artemis Program, everything goes smoothly. We will see some flavor of Kubernetes and containers running on the lunar center.

Luke Schantz: That is so cool.

Joe Sepi: Yeah.

Naeem Altaf: That won't be... For example, it will be edge from elsewhere. I think it's relative. From there, it'll be they call it on- prem.

Luke Schantz: That's a good point you bring up, too, because I feel like the term... I think that's why edge is such a confusing term sometimes-

Naeem Altaf: It is.

Luke Schantz: ...because one person's edge is another person's regional CDN or something. It's all relative to where your work's being done.

Naeem Altaf: Yeah, the spectrum is huge. Let's say all the way from traditional ways of where on- prem, you have data centers, then the evolution in cloud came into the picture. Then, you have these... Let's say if you talk about telco. They have NDS, or the switching officers. You can just... Wherever you stop you can say, " This is my edge." You stop there. If you go one more further, you say, " Now, that cell site," which if you have looked at the cell site, they have that room next to it. It has computers and everything. You can say, " That's my edge." Or you can say, " This Raspberry Pi in my hand, which is talking to the seller network all the way back to the cloud, this is my edge." Or somebody can say, " My sensor is my edge." Spectrum is huge for edge computing, yes.

Joe Sepi: That's really interesting. I want to ask you what you are excited about in the future, but I feel like a lot of what we've been talking about is what you're excited about and coming up. Another project I think we've talked about, I don't know if we want to get too much into it here, but the Mayflower Project, are you involved in that as well?

Naeem Altaf: Yeah, so me and Mike, Eric, and Brett and Don were actually from the Mayflower. We actually worked together three years ago on a napkin to put this thing together. " How are we going to build this whole thing?" Yes, and so we were involved from the very first day and you have seen last year it went to the waters and most of it's happening. To me, what Mayflower is, it's very simple in words. It's a lab. It's a lab in the ocean, and folks, if you can go to mas400.com, you will see the live dashboard, everything. That's what our team, folks, our team they built. I want to give a shout-out to Ben inaudible and my team. He did all of the inaudible of that, most of the back end stuff. What we see, this boat is a lab. Think about this. I was actually just talking to Brett two days ago. I was like, "Okay, you know what? Your lab is floating in some far, far areas in the ocean because almost one- fourth of the world is oceans. Plenty of water there, and there is no connectivity. What if we had our satellite and I want to say because of GPS I know where you are, can you go and check for certain things around here? We have so many sensors there and I send you a payload. I send you a containerized code relayed to the satellite, and come to your board. Then, you can do your stuff and send me the results back." That's the next thing that which we are thinking about to connect space to the ocean and to the board with the satellite and how we can do communication. That is edge. That's an edge in the ocean somewhere and we have an edge in the space, so we are trying to bring all these things together now.

Joe Sepi: Yeah, that's fascinating. Really cool stuff.

Luke Schantz: Reminds me, too, maybe two years ago, Joe and I were doing an event in New York at a makerspace called fat cat FAB LAB, and one of our advocates at the time bought one of those Iridium Satellite RockBLOCK development boards. It was this... He did the demo, sent it out through like the cloud, and then everyone's looking out the window. It was like a lot of anticipation. Is this going to work? Is it going to come down? It did. We ended up... We sent the message through the cloud and we got it back through the thing and it was like a huge hit, but I imagine the bandwidth there was limited and the coverage was somewhat limited, but now we're seeing this. You're mentioning there's going to be over this next few years a proliferation of global bandwidth connectivity with, like you said, at least three or four different companies or governments putting up these huge communication networks.

Naeem Altaf: Yes, and for anybody out there who's listening, if you're interested just to start playing with this stuff, you need a parabolic dish, which is like a hundred, 130 bucks, and then a Raspberry Pi and a SDR < software defined radio. Just plug that in, and if you have seen the images from the GOES 16 and 17 satellites, the full globe picture with the weather pattern, everything, the colorful pictures, you can get that every 15 minutes. I have in my back yard, and because it's not geo, so you point to it, that's it. Every 15 minutes you will get the latest picture which everybody's getting. You will get in your inaudible it's that easy, and you will learn a lot how the communications are happening between the satellite because you have built a ground station now. You have a ground station at the back of your home, and you plugged into your laptop and you're getting these images. If anybody does, just Google it, " GOES 16 open source SDR." You will get the whole stuff written there. Just follow the instructions, you will have it.

Speaker 3: Yeah, that's really cool. That's... I've mentioned before, I have a young son. I'd love to put that together and just really explore the connection between us and what's out there in space and whatnot. It's really cool. I was going to ask, too, I'm glad you jumped in with that, but are there other ways to get general, regular people involved in some of this work in that sort of tangible way?

Naeem Altaf: Yes, so there's a community called Sat Knocks, and basically these are amateurs all over the country and just all over the world. You can get the core inaudible Raspberry Pi and they're basically getting signals. You just need a software to find where you dongle. It's a USB device, and you can start getting signals from... They're actually the first ones whenever you launch the CubeSats or small sats, they basically broadcast a message to all these amateurs who can find first signal, and it's very cool. If you want to learn about that, yes, be part of that community.

Joe Sepi: Very cool.

Luke Schantz: We should do that. I feel like that would be a great project. We could even... I'm sure, like you said, there's tutorials out there, but maybe we could even-

Naeem Altaf: Yes.

Luke Schantz: ... doa blog post on IBM Developer about that and, yeah, that's so interesting. I definitely want in.

Joe Sepi: Yeah, we had a dish satellite on our new house here, but we use regular DSL or whatever it is, and so I actually took the dish off and we're using the arm part to hang a bird feeder inaudible-

Naeem Altaf: I did the same when I got that parabolic dish. I used an old arm of that dish and I plug it there and put it in the floor and it's now pointed up, yeah.

Joe Sepi: Great. I'm glad I didn't take the whole thing down. I can just hook up a better parabolic dish on there, too, and just get started. That's great.

Luke Schantz: I also wanted to mentioned that regarding the CubeSat stuff, I found a blog post that your team put up and that's the link to the blog post that has some stuff about the CubeSat Project.

Joe Sepi: Cool, and it looks like we've got a question from Facebook. " What's the role of Kubernetes in this from edge to satellite, question mark?"

Naeem Altaf: Yes, so this was the experimentation which we did with the ISS for the DNA sequencing. Basically, we containerize the code and we use the queue platform so we can run multiple workloads in the future, but again, depending upon the footprint of your hardware, let's say if you only have Raspberry Pi, you can use PortMap to run your container. The goal was to run a containerized port, which is a self- containable clean. I don't have to be dependent upon because when you're in the space there, it's air gap environment. You don't have luxury to go and let me go and grab something from the internet. You can't do that. It has to be everything enclosed, so that was the reason. Plus, we had higher resources to run this DNA sequencing, but for example, if I had to my CubeSat, unless I don't have a very lightweight footprint of queue, I will probably use inaudible.

Joe Sepi: Interesting, and you... I don't know if it's what we've mentioned here or in the video I saw Red Hat as well. Is that OpenShift running on some of this?

Naeem Altaf: Yes.

Joe Sepi: Is it the code- ready containers? I'm curious, do those have... The code- ready containers, are they able to be deployed on these smaller systems? I was under the impression that you needed a development machine to really run those, but how is that working?

Naeem Altaf: Yes. Yeah, because, again, on this system, these are like the enterprise class servers, so it wasn't a problem-

Joe Sepi: Okay.

Naeem Altaf: ...to run this, but if we are looking at a lightweight, then yes. You need a little bit more power, but I know there's some work been happening, but you'll see some interesting stuff happening in next few months.

Joe Sepi: Yeah, really cool, really cool. Okay.

Luke Schantz: Just... This is like a question/ comment. Joe, I know you're deep into the JavaScript, but if you do any computing for these satellites, no hot linking any MPM stuff. You got to put that in the container. It's just not going to-

Joe Sepi: Yeah.

Luke Schantz: ...can't have that.

Joe Sepi: Yeah.

Luke Schantz: Sorry, I-

Joe Sepi: This is dumb. But in last year's OpenJS World Conference we had a NASA astronaut as one of the keynote speakers and, of course, my question is MPM modules are heavier, the heaviest thing in the world. Yeah, I don't know if that's actually true, but I had to ask. Silly. Cool. Are there any other things that you are excited about that you want to share, Naeem, before we look to wrap up?

Naeem Altaf: Yes, a couple of more things. One thing as IBM's... We are going to very soon release an official SpaceTech report, our vision. We reached out to industry leaders from... Instead of just going to a hitting NASA, ESA 30 times, we went to all the different startups on the terrestrial networks and see what they think of space. How are they related to the space? To get a much broader view of what the industry thinks about space. We will have our vision, what we think. We have a long history, like you mentioned in the beginning, so we are no strangers to space. We put out our vision there based on current and future technologies, and then we will have a point of view from all these industry leaders. Stay tuned, that will be coming soon. Just a teaser, I think I mentioned about the CubeSat stuff. I want IBM to be seen as democratizing access to space for everyone, that space because imagine to build something and launch, it's a privileged thing. It's only a few nations, and within those three nations, only few groups can do that. Think about majority as well because space is an inspiration. When you talk about space to anyone, regardless of whatever field they are in, it's very inspiring. It's just like we look at above every night. Every night I just go up and then just look at the sky. It's so peaceful and calm and, sorry, it gives you inspiration. The goal over here is that we open it up to everyone in the world, that what the goal area is with kids sitting somewhere in far, far part of the world and they write this very simple Python code and just submit to us, and we will push that code all the way to the space. It'll do some computation and come with the results, so they can also feel attached to the space. Yeah, that's the quote.

Joe Sepi: Yeah, that's cool. If you could put a message up in space, that would be even cooler, but I know that's probably... but it's true. I walk the dogs every night and I do it. I live in the woods now and the lights are out and I just stare up at space and it's just really amazing. It's really cool.

Naeem Altaf: Yes.

Luke Schantz: There was some company a while back I saw that was doing some sort of artificial meteor showers where they would basically send up some sand and then throw it at the certain time and you'd see these thing, but I don't know if that really... That didn't scale I'm pretty sure. Yeah.

Joe Sepi: It's a good question, though. Is that part of the air traffic controller work that's being done? Not just are they colliding, but any sort of meteors or meteorites coming into the atmosphere?

Naeem Altaf: Any activity over there, yes, because they have the equipment to watch for those things. Yes, definitely.

Joe Sepi: Yeah.

Naeem Altaf: Measure it and share it. Sharing is key, yeah.

Joe Sepi: Yeah, yeah, and are you finding in this work collaborating with other entities and governments... Like I said earlier, people's the hard part, but are people generally working together well and coming up with policies collaboratively in a positive way?

Naeem Altaf: Yes, so one of the leading professors, his name is Moribah Jah. He's from UT Austin here, and he is like the most vocal in this space and he's very well... works with the knowledge and everything. We have been collaborating with him since last year in these open source projects for the space. We were working with him. I actually got a call with him yesterday also. We are going to look for the space. Weather has an impact on the objects in the orbit. Does it move up? Whatever happens, the characteristics of the object. Yes, we are collaborating with the universities very closely. Yeah.

Joe Sepi: That's great. That's great. I encourage folks to follow Naeem on Twitter and the work that he's doing because there's just so much interesting stuff happening now and happening in the future. It's really cool.

Naeem Altaf: Thank you.

Luke Schantz: Yeah, and it seems like if we get this figured out and we do this, the potential upside here is we live on this finite resources, and if we can make this transition and get into space mining, get communication, energy, it's really it could be the new boom up there.

Joe Sepi: Yep. Very cool.

Luke Schantz: Let's see, if there's any other questions coming through. I think we are going to try to wrap at the top of the hour because that's what we're scheduled for. There was some question about... It's not really a question, but it says something about making the environment safe for... I think it's a question about, is the DNA... does it get damaged in space? Or, I guess, I imagine if it's inside the space capsule that's radiation- shielded.

Naeem Altaf: Yes, it is.

Luke Schantz: Yeah.

Joe Sepi: It seems like we have another question here. " Can satellite images provide real- time traffic control maybe for self- driving communications, too?" Do we have a system perhaps good enough to continuously track so many areas and so many vehicles, I guess, maybe is the question?

Naeem Altaf: Of course, GPS is there, so if you use waves and all these apps about the traffic stuff, but the interesting thing, which we just heard two days ago that Starlink is going to open up for the mobility for the products, vehicles, ships. I think that the more productivity you have, the more precision will come in the picture. Yeah, I think this is a big disruptor. For example, for the people who are listening, think about these use cases. Think about when the natural disaster, anything occurs like that, or the fires or any of these things for the first responders. You're in very remote areas. If you have little bit of light compute and you have this Starlink antenna, you just plug it there. Now, you're connect with the world. There are so many use cases because interconnectivity is not everywhere. Even though we assume that maybe it's everywhere, but if we go a little bit north of our cities, the connection drops drastically, the speed and everything. That's why the FCC awarded these big contracts to provide a broadband connectivity to all over the U. S. That's where the governments are coming up now and thinking, " Okay, if I will have a connectivity at 2, 300 meg in far, far areas, what can I do?" You can do so much environmental monitoring, so many things. Yeah, I think it's-

Joe Sepi: inaudible.

Naeem Altaf: ...a game changer. Let's keep an eye on that.

Joe Sepi: Yeah, really cool.

Luke Schantz: Thank you so much for your time. It's been really a pleasure having you as a guest, and thank you for our audience for spending this time with us chatting. I think, Naeem, let's maybe later on this year at least come back, give us up inaudible because there's much going on in this space. We'd love to have you back on.

Naeem Altaf: Thank you so much, Luke and Joe, for having me on. It's a pleasure.

Joe Sepi: Yeah. It's been great.