Interview with a popularizer of astronautics, ballistic programmer Anton Gromov
A series of interviews with DUMP conference speakers is continued by a conversation with Anton Gromov, a ballistic programmer at Dauria Aerospace, the author of the Sea of Clarity project, a constant host of SpaceX broadcasts, a popularizer of astronautics. Anton spoke about satellite programming, the state of private space exploration in Russia, and what kind of work there is for programmers in space exploration.
Anton, you have an unusual position - a ballistic programmer. Please tell us in more detail what ballistic programmers do. What projects are you currently working on?
Honestly, I don’t know any more people whose job title is called that way :) Basically, I am currently programming on-board software for our next satellite. In addition to this, sometimes I solve ballistic problems when they appear, but only, so to speak, of low and medium complexity, because my main task is programming.
You work for the private space company Dauria Aerospace. How did you get there? How is everything at Dauria Aerospace? How is your work organized?
Do not believe me, I sent a resume and went through an interview! By that time, I already knew a couple of people from the company on the project of the lunar satellite , in which I was engaged in ballistics. There, my task was to calculate a stable low orbit near the moon.
Compared to other companies in the space industry, we have a fairly young team. The backbone of the team was formed 6-7 years ago from industry experts. As for my work, the tasks, of course, are set very clearly, but there is great scope for taking the initiative and expanding the area of responsibility, which I really like. The company is still small, and this compares favorably with the ossified enterprises of our space industry.
What technologies does Dauria Aerospace use? Which stack? What special tools do you use?
Oh, there are an order of magnitude more technologies and tools than in IT, where I come from. On the satellite it is necessary to calculate the design, circuitry, strength, magnetism, temperature interactions, ballistics, radio, optics. For all this, of course, their own tools. As for pure on-board programming, we restrict ourselves to the C + C POSIX library + gcc. For ballistics, I mostly use Python, sometimes in combination with the scripting language of NASA's GMAT package.
Are there bugs in production in space? If not, how do you achieve this quality? If so, then what happens?
Everywhere there are bugs in production :( One of the famous examples is the first launch of the European Ariane 5 rocket. She used part of the control system from the previous Ariane 4 rocket. In one place, the variable responsible for horizontal acceleration was transferred from double (64 bit) to short int (16 bit) On Ariane 4, horizontal acceleration fit into both types of frames, so there were no range checks when casting.
I must say that this was not the forgetfulness or carelessness of the developers, but the optimization of the effectiveness of the code, which in space technology has to work on much slower hardware. But the Ariane 5 was designed differently, it had more acceleration, and at the 39th second of the flight it no longer fit into 16 integer bits. An exception arose that was not adequately handled, the rocket began to sharply change its orientation and was destroyed by aerodynamic forces. For some reason, tests of this part of the control system were not carried out in combination with other systems, otherwise the error would have been identified in advance.
Another, probably more famous example was the recent launch from Vostochny. The headings were "mixed up by cosmodromes", but in fact, if I understood correctly, the control system code simply did not check which direction to turn (instead of turning by 2 degrees, a turn of 362 began to be performed). At other cosmodromes, this rotation worked as it should, and before launching from Vostochny, again, they did not conduct comprehensive tests of all the systems together, they checked only individual systems or individual links.
In addition, two years ago, the Japanese lost a cool orbital telescope due to a software error, and, of course, the list goes on. This is not entirely programming, but one cannot fail to mention the loss of the American Martian satellite Mars Climate Orbiter due to the fact that Lockheed Martin transmitted data to control it in feet and inches, while NASA has long used the metric system.
In a word, in the vast majority of cases, problems arise precisely at the junction of different areas of responsibility. And within the framework of one system, intensive tests help to avoid errors.
If not a secret, then what are the salaries of programmers in the space sector?
Various. On average, the hospital average is probably smaller than in IT, and there isn’t much freelance here.
What work is there for programmers in space? What kind of specialists do the space industry need now?
If we are talking about Russia, in addition to our company, now we are making a new manned ship, new missiles (Angara and Soyuz 5), new satellites, lunar landing stations, the Martian landing platform for ExoMars. For all of these systems, both on-board and ground-based software are needed, so overall there is something to program. In addition, there is a need for testers who are familiar with the specifics of space software, because often the same teams that write the code are involved in the tests.
You have designed, assembled and launched into orbit three private satellites. Have you been involved in their development? Tell me more about this?
No, I’ve been in Dauria for only six months, but I am actively adopting the experience gained by the company in these projects. Unfortunately, the last two devices could not be tested in flight due to an emergency situation when putting it into orbit. Nevertheless, during the development and creation of the apparatus, of course, they did a great job, the results and experience of which can be used further.
From the side it seems that Russian private space exploration is rather difficult to develop. This is true? In your opinion, what hinders the development of private astronautics in Russia? Will we have our own SpaceX?
It can be difficult, yes. Here I’d better just let you read the article "Roskosmos" against private traders: where 300 million flew ", from it as a whole everything is clear. The main problem is that almost any private space company is perceived as a competitor to Roscosmos, which already has all the necessary competencies without it.
As for SpaceX: in this, the United States, on the contrary, relied on private traders, thus spurring the market and allowing it to develop. Anyway, the first steps and the first significant investments in SpaceX were made by private individuals anyway, and when they showed NASA with their money that they could, they were given a contract.
What to expect from private space companies in the near future? What projects do you plan to launch?
This year we plan to launch our satellite Aurig, he will photograph the Earth at a resolution of up to 2.8 meters per pixel. This is a very good indicator for such sizes and masses: a satellite with about a microwave. This will be a demonstration mission, after which it will be possible to create a whole group of such devices. In addition, we are actively working on the project of the Atom geostationary platform.
If about other companies, then in Russia there are a couple of start-ups with the goal of building a light rocket for the international market. While they are at a relatively early stage, but I hope they will develop.
In the west, private space exploration is on a huge rise. In addition to several rocket startups with significant success, there are many exclusively satellite manufacturing companies, including Planet, which recently announced that it photographs the entire surface of the land every day. There are OneWeb and SpaceX, which for a couple want to launch 15,000 satellites of the orbital Internet, and many others.
You are one of the popularizers of astronautics in Russia. How and why did you start doing this?
Because I love space and love to talk about it, everything is quite simple here. I took up the ballistics of the lunar satellite, then gave a lecture on ballistics, then a few more lectures. At the end of 2015, I began to broadcast Falcon 9 rocket launches. It was just my first broadcast that fell on the first successful landing of a stage in history. There were many people who were inspired by the successes of SpaceX, and not all of them know English, so I still conduct these broadcasts. And in the summer of 2017, I launched the cognitive channel Sea of Clarity on YouTube, this is now my main popularization project. Unfortunately, it is not possible to devote as much time to it as we would like, but there are big plans.
Please tell us more about your projects? Do you feel that interest in this topic is growing?
In addition to working in Dauria, I’m doing the aforementioned channel on YouTube, I’m conducting the aforementioned Russian-language broadcasts of SpaceX launches , periodically giving lectures, in charge of ballistics in the lunar satellite project, which I also spoke about. By the way, now we are just completing the next iteration of the preliminary project. A couple of projects so far are only in the plans or at the stage of evaluating the idea.
Perhaps one more project of mine will be interesting to the Habr audience - the exact model / simulator of the Solar System on JS. It will soon make available all the ever running devices (more than 40,000), you can view it all, create your own trajectories, play with them, and a lot more plans. The project is being done exclusively on a volunteer basis, and now it has not even reached the alpha version stage, but is available at urbain.ru . If someone is interested in helping, here you can study the list of plans, but you can write to me in telegrams.
As for the public interest in the topic of space, it is quite difficult for me to evaluate it, because my whole life consists only of it. I am glad that space projects at crowdfunding collect normal amounts (the latterthe project is currently ending on March 29).
The other day I was traveling in a space carriage in the Moscow subway, and a month later the famous Cosmos pavilion at VDNH should open. A new large planetarium has opened in St. Petersburg, and the regions are also slowly developing the advancement of space into the masses, which is good.
On April 13 in Yekaterinburg, Anton will speak at the DUMP conference in the Science section. Anton will tell you which satellites will be launched in the near future, who will produce them, and which rockets and technologies are being developed to launch them into orbit.
Thanks to our partners who make the conference possible: the general partner - Sberbank-Technology , the conference partners: SKB Kontur , Naumen , Tinkoff.ru , ProSoft Sistems , Very Interesting , Skb_Lab .
Anton, you have an unusual position - a ballistic programmer. Please tell us in more detail what ballistic programmers do. What projects are you currently working on?
Honestly, I don’t know any more people whose job title is called that way :) Basically, I am currently programming on-board software for our next satellite. In addition to this, sometimes I solve ballistic problems when they appear, but only, so to speak, of low and medium complexity, because my main task is programming.
You work for the private space company Dauria Aerospace. How did you get there? How is everything at Dauria Aerospace? How is your work organized?
Do not believe me, I sent a resume and went through an interview! By that time, I already knew a couple of people from the company on the project of the lunar satellite , in which I was engaged in ballistics. There, my task was to calculate a stable low orbit near the moon.
Compared to other companies in the space industry, we have a fairly young team. The backbone of the team was formed 6-7 years ago from industry experts. As for my work, the tasks, of course, are set very clearly, but there is great scope for taking the initiative and expanding the area of responsibility, which I really like. The company is still small, and this compares favorably with the ossified enterprises of our space industry.
What technologies does Dauria Aerospace use? Which stack? What special tools do you use?
Oh, there are an order of magnitude more technologies and tools than in IT, where I come from. On the satellite it is necessary to calculate the design, circuitry, strength, magnetism, temperature interactions, ballistics, radio, optics. For all this, of course, their own tools. As for pure on-board programming, we restrict ourselves to the C + C POSIX library + gcc. For ballistics, I mostly use Python, sometimes in combination with the scripting language of NASA's GMAT package.
Are there bugs in production in space? If not, how do you achieve this quality? If so, then what happens?
Everywhere there are bugs in production :( One of the famous examples is the first launch of the European Ariane 5 rocket. She used part of the control system from the previous Ariane 4 rocket. In one place, the variable responsible for horizontal acceleration was transferred from double (64 bit) to short int (16 bit) On Ariane 4, horizontal acceleration fit into both types of frames, so there were no range checks when casting.
I must say that this was not the forgetfulness or carelessness of the developers, but the optimization of the effectiveness of the code, which in space technology has to work on much slower hardware. But the Ariane 5 was designed differently, it had more acceleration, and at the 39th second of the flight it no longer fit into 16 integer bits. An exception arose that was not adequately handled, the rocket began to sharply change its orientation and was destroyed by aerodynamic forces. For some reason, tests of this part of the control system were not carried out in combination with other systems, otherwise the error would have been identified in advance.
Another, probably more famous example was the recent launch from Vostochny. The headings were "mixed up by cosmodromes", but in fact, if I understood correctly, the control system code simply did not check which direction to turn (instead of turning by 2 degrees, a turn of 362 began to be performed). At other cosmodromes, this rotation worked as it should, and before launching from Vostochny, again, they did not conduct comprehensive tests of all the systems together, they checked only individual systems or individual links.
In addition, two years ago, the Japanese lost a cool orbital telescope due to a software error, and, of course, the list goes on. This is not entirely programming, but one cannot fail to mention the loss of the American Martian satellite Mars Climate Orbiter due to the fact that Lockheed Martin transmitted data to control it in feet and inches, while NASA has long used the metric system.
In a word, in the vast majority of cases, problems arise precisely at the junction of different areas of responsibility. And within the framework of one system, intensive tests help to avoid errors.
If not a secret, then what are the salaries of programmers in the space sector?
Various. On average, the hospital average is probably smaller than in IT, and there isn’t much freelance here.
What work is there for programmers in space? What kind of specialists do the space industry need now?
If we are talking about Russia, in addition to our company, now we are making a new manned ship, new missiles (Angara and Soyuz 5), new satellites, lunar landing stations, the Martian landing platform for ExoMars. For all of these systems, both on-board and ground-based software are needed, so overall there is something to program. In addition, there is a need for testers who are familiar with the specifics of space software, because often the same teams that write the code are involved in the tests.
You have designed, assembled and launched into orbit three private satellites. Have you been involved in their development? Tell me more about this?
No, I’ve been in Dauria for only six months, but I am actively adopting the experience gained by the company in these projects. Unfortunately, the last two devices could not be tested in flight due to an emergency situation when putting it into orbit. Nevertheless, during the development and creation of the apparatus, of course, they did a great job, the results and experience of which can be used further.
From the side it seems that Russian private space exploration is rather difficult to develop. This is true? In your opinion, what hinders the development of private astronautics in Russia? Will we have our own SpaceX?
It can be difficult, yes. Here I’d better just let you read the article "Roskosmos" against private traders: where 300 million flew ", from it as a whole everything is clear. The main problem is that almost any private space company is perceived as a competitor to Roscosmos, which already has all the necessary competencies without it.
As for SpaceX: in this, the United States, on the contrary, relied on private traders, thus spurring the market and allowing it to develop. Anyway, the first steps and the first significant investments in SpaceX were made by private individuals anyway, and when they showed NASA with their money that they could, they were given a contract.
What to expect from private space companies in the near future? What projects do you plan to launch?
This year we plan to launch our satellite Aurig, he will photograph the Earth at a resolution of up to 2.8 meters per pixel. This is a very good indicator for such sizes and masses: a satellite with about a microwave. This will be a demonstration mission, after which it will be possible to create a whole group of such devices. In addition, we are actively working on the project of the Atom geostationary platform.
If about other companies, then in Russia there are a couple of start-ups with the goal of building a light rocket for the international market. While they are at a relatively early stage, but I hope they will develop.
In the west, private space exploration is on a huge rise. In addition to several rocket startups with significant success, there are many exclusively satellite manufacturing companies, including Planet, which recently announced that it photographs the entire surface of the land every day. There are OneWeb and SpaceX, which for a couple want to launch 15,000 satellites of the orbital Internet, and many others.
You are one of the popularizers of astronautics in Russia. How and why did you start doing this?
Because I love space and love to talk about it, everything is quite simple here. I took up the ballistics of the lunar satellite, then gave a lecture on ballistics, then a few more lectures. At the end of 2015, I began to broadcast Falcon 9 rocket launches. It was just my first broadcast that fell on the first successful landing of a stage in history. There were many people who were inspired by the successes of SpaceX, and not all of them know English, so I still conduct these broadcasts. And in the summer of 2017, I launched the cognitive channel Sea of Clarity on YouTube, this is now my main popularization project. Unfortunately, it is not possible to devote as much time to it as we would like, but there are big plans.
Please tell us more about your projects? Do you feel that interest in this topic is growing?
In addition to working in Dauria, I’m doing the aforementioned channel on YouTube, I’m conducting the aforementioned Russian-language broadcasts of SpaceX launches , periodically giving lectures, in charge of ballistics in the lunar satellite project, which I also spoke about. By the way, now we are just completing the next iteration of the preliminary project. A couple of projects so far are only in the plans or at the stage of evaluating the idea.
Perhaps one more project of mine will be interesting to the Habr audience - the exact model / simulator of the Solar System on JS. It will soon make available all the ever running devices (more than 40,000), you can view it all, create your own trajectories, play with them, and a lot more plans. The project is being done exclusively on a volunteer basis, and now it has not even reached the alpha version stage, but is available at urbain.ru . If someone is interested in helping, here you can study the list of plans, but you can write to me in telegrams.
As for the public interest in the topic of space, it is quite difficult for me to evaluate it, because my whole life consists only of it. I am glad that space projects at crowdfunding collect normal amounts (the latterthe project is currently ending on March 29).
The other day I was traveling in a space carriage in the Moscow subway, and a month later the famous Cosmos pavilion at VDNH should open. A new large planetarium has opened in St. Petersburg, and the regions are also slowly developing the advancement of space into the masses, which is good.
On April 13 in Yekaterinburg, Anton will speak at the DUMP conference in the Science section. Anton will tell you which satellites will be launched in the near future, who will produce them, and which rockets and technologies are being developed to launch them into orbit.
Thanks to our partners who make the conference possible: the general partner - Sberbank-Technology , the conference partners: SKB Kontur , Naumen , Tinkoff.ru , ProSoft Sistems , Very Interesting , Skb_Lab .