The Nextspace™ Broadcast

Episode #003 Building satellites against cyber attacks with Dmitry Mikhaylov

[Rachana, Host]
Hello and welcome to the NextSpace Broadcast, brought to you by Reflex Aerospace. As a satellite manufacturer operating on the upstream end of the space value chain, this podcast series is part of our deep engagement efforts with current and potential end users of space applications to better understand their requirements and to develop smarter solutions for the future together.

I'm your host, Rachana Mamidi. And today with us is Dr. Dmitry Mikhaylov. at the National University of Singapore, and he is a graduate of the UK's University of Warwick and he holds a Ph.D. in artificial intelligence. He is the author of numerous patents and publications and books in multiple languages, and he has an extensive experience in applying satellite imagery.

And he also uses AI and big data concepts to engineer new products for the cybersecurity market. And given Dr. Mikhail's incredibly extensive experience around diverse topics which are so interesting for us today, let's focus on two of these topics. First one, let's dive into cybersecurity with increasing satellite connectivity.

So what cybersecurity means when there is increased satellite connectivity, how can we build more resilient satellites that are resilient to these cyber effects and now with the onset of quantum cyber attacks. So let's see how we can build more resilient systems. Welcome to the podcast Dr. Mikhaylov.

[Dr. Dmitry Mikhaylov]
I'm actually focused in AI and big data area. And what is quite important from my point of view is that we start this talk not from the technologies but from the aspect of cybersecurity and I'll tell you why, because I've seen a lot of solutions that try to penetrate the market, do something very fast, very efficiently and provide a new product.

And they don't think about cybersecurity at all. And then at the end of the day, they start patching. That is not a good approach to cybersecurity. They start to understand that they have solutions so vulnerable, and especially nowadays, you can lose a whole business. So instead of speaking about cybersecurity as a lost topic, it would be good to start with it because it's a crucial thing.

And nowadays, especially in aerospace, in satellites, because satellites are not regulated that well, space belongs to everybody and there are no regulations. Same as sea, same space. United Nations, I think, have to make much more efforts in regulations of space because nowadays I'll give you an example. I can make a start up on launching a small satellite into space. And this satellite, as far as it's a small startup, will not be so secure. So anybody can actually hack it. And what happens afterwards, anybody who hacks it can target it to what's a bigger satellite and make a major collision. And nobody thinks about it. Nobody speaks about it.

Satellites are so close to each other. Then even one small satellite from a startup can cause a huge damage and there is not actually any kind of regulations on their cryptography. On the cybersecurity side, everybody wants to make a more cost effective satellite, and that's a huge problem and a big challenge in the future.

[Rachana, Host]
Yeah, absolutely. While building satellites, cybersecurity considerations are more of an afterthought than which go in at the beginning of, you know, designing the satellite. So I absolutely agree with that. But first of all, let's perhaps start with the basics. How are cybersecurity vulnerabilities quantified in the first place?

[Dr. Dmitry Mikhaylov]
If we speak about satellites, there are problems like GPS jamming - it's when you try to jam the signal of a satellite or not only GPS but just jamming the signal, eavesdropping - if you don't have a very good cybersecurity, cryptography protection, spoofing and hijacking. can take control over the satellite because you didn't do quite a good protection there.

And to make a very good satellite protection, you have to think out first, who can do damage to your satellite? Who are those hacktivists who can do that and what is their motive? This is very important. Of course, most of the satellite companies say, well, we don't have that many enemies, that we are not vulnerable.

Nobody will think about taking down the satellite that's just tracking, say, carbon credits, whose interest is that? But it's not the right approach because at the end of the day, somebody who wants to take or hijack your satellite may not be even thinking about your company. They can think about moving the satellite in the wrong orbit or wrong position and target a military satellite, for example. So this rejection of being of non-interest to any hackers is not the right approach.

[Rachana, Host]
That’s very interesting. So when we're trying to make satellites more resilient to cyber attacks, can the methods of IoT malware detection, for example, be applied to satellites because they share these similarities, right? Iot devices and satellites have these similarities that both of them are mostly remotely accessible. And they also have these tighter, let's say, on board or on chip data handling protocols. So could this be one method to mitigate cyber attacks?

[Dr. Dmitry Mikhaylov]
Partially, yes.. But at the end of the day, many satellite companies forget about checking their hardware and software of a satellite just before the launch. They just skip this scanning and proper audit. And honestly speaking, if you even try to Google any service that makes an audit of a satellite equipment, you will not find a single company.

And it's a huge problem because nobody is making the audit of the whole hardware and software before the launch. After the launch, of course, nobody cannot get access to the satellite in space, if your channel is properly encrypted. But this encryption is also usually not really cared about. Why? Because it takes money. It takes energy from the satellite.

Encrypting and decrypting something is an additional procedure, and most of the companies try to skip it. When they try to skip it, they can have hackers penetrating this channel, trying to take control of the satellite. So it's number two. The first is audit and the second is proper encryption. And the third one, it is also very important if you have on land, software and hardware that controls the satellites, most of the companies just neglect that this software can be attacked.

And then through the channel, even if it's encrypted, the satellite can be attacked. So we have three components and most of the components unfortunately are not taken care of.

[Rachana, Host]
So what do you think would be the ideal way to design, develop and build a satellite to make it completely resilient or perhaps if not completely as much as possible, resilient to cyber attacks?

[Dr. Dmitry Mikhaylov]
You know, unfortunately and I have been in this area for like ten years, maybe more. We've been introduced with smartphones then with industrial control systems, and IoT and all these new innovations. For some years, they had a lot of hackers trying to take benefits from unprotected assets and then regulations were imposed.

And cybersecurity companies, after the regulations, started to develop the product. Why they do it after these regulations are imposed is because regulations create a market for them. That means they don't have to sell by running after every company and selling their product. That means there's a company out trying to buy the product that helps them to be certified according to the cybersecurity regulations. Unfortunately, in this industry, in the satellite industry, there are no regulations so far, or at least the regulations are not that good, not that strict. And my strong belief is that until we have strong regulations in and not on the government level, it has to a United Nations (UN) kind of regulation. So anybody who launches something into space will have to go through some procedure.

Only in this case, we will have a really cyber secure space. Before that, I don't think it will work because companies will still try to save money on this .

[Rachana, Host]
I agree with you. The satellite market is not really the most regulated compared to the other sectors.

[Dr. Dmitry Mikhaylov]
Insurance companies can also take the make a step because most of the satellites now are insured. But again, we don't see any move from insurance companies so that they make this cybersecurity obligatory. The service of audit obligatory, for example.

That would actually be a no brainer because insurance company is the biggest stakeholder at the end of the day of money wise, financials wise. Hopefully someday. So we are talking about cybersecurity and this brings into the picture - satcom networks. Shared satellite networks are becoming more and more popular. And what is the magnitude of risk for, let's say, an end user, not just the satellite, but for let's say I'm an end user who's connected to a SATCOM network?

What is the magnitude of risk to me when I'm using a SATCOM network that is run on shared satellite networks compared to a self-owned and a self-operated satellite network?

There is not right answer here because if you make your own satellite and you are on your own and you have very good cyber secure protection, it's a very good solution. If you don't have that and you have a shared network, it can be a better solution because a shared network can also have a higher, higher protection level.

Of course, the best solution, if you have your own network and your cybersecurity, then you have enough money to guarantee cybersecurity of the solution. So you have to always understand basically who is this threat actor, who is a hacker, who can hack you, who wants to hack you and act according to this? Because if it's a foreign intelligence, it's one case.

If it's a student hacker, it's another case. If it’s a competitor, then it’s another case . So you have to understand who is a hacker who can hack you, what he has at hand and what are his intentions and incentives. That is the key question. And based on that, you can make your choice.

[Rachana, Host]
This is a very useful insight, Dr. Mikhailov. So, at the end of the day, whatever solutions we are going to implement towards resilience, towards making the satellite or the SATCOM network resilient against cyber attacks, is basically figuring out the source of the potential threats. So you mean to say that every single solution would basically arise out of identifying the source of this threat?

[Dr. Dmitry Mikhaylov]
Of course, because, for example, you can make a good cryptography encryption for the channel, right? But if your opponent, a big country with very good supercomputers and with great hackers and they really want to put your satellite down, you really can do nothing against it. Or you have to invest so much into cybersecurity that your business will not be affordable. You have to have this model in hand before you make a choice.

[Rachana, Host]
That’s actually good to know. So that means there is no blanket solution against cybersecurity and it's all going to be just like on the ground, for terrestrial applications - it's going to be levels of cybersecurity.

[Dr. Dmitry Mikhaylov]
Absolutely.

[Rachana, Host]
These are very, very interesting insights. Thank you very much for that. Hopefully, we'll try to incorporate all these useful pointers into our satellite designs and hopefully add an additional layer of assurance to our potential customers.

Having discussed all that about cybersecurity, I know it can be a bit daunting, but yeah, let’s have our fingers crossed and hopefully the whole satellite industry comes up with better solutions and better mitigation methods against cyber attacks. And let's move on to the next topic, which is also quite exciting.

Dr. Mikhaylov has also done extensive work in verification of carbon credits using satellite imagery. And we can perhaps now talk about the role of satellite services in general in the whole carbon regulation market. So first of all, Dr. Mikhailov, is satellite imagery currently being used for carbon credit verification?

[Dr. Dmitry Mikhaylov]
Yes, they are used, but at this stage at least of the development of this market, I would focus on methane because it has actually 22 times greater global warming potential than carbon. And it's easier to track from satellites. So at least my experience was focused within the last couple of years on the destruction caused by methane emissions, and they are quite easy to track from satellites. So I can tell that satellite imaging is already playing a great role, at least in making these methane impacts lesser.

[Rachana, Host]
During your research or your research work so far, have you encountered any barriers to accessing satellite imagery? Well.

[Dr. Dmitry Mikhaylov]
Well, yes. It’s a big problem nowadays because as we all know, there are several systems of satellites, passive systems that use the sun as a source of electromagnetic radiation and active remote sensing systems that carry on-board their own electromagnetic radiation sources. And there is a lack of images and there is a lack of data that is accessible, especially free data. In most cases, you have to pay quite a lot to get access to this information. And most of the companies who try to detect this carbon offsets, they try to save money, of course. Because most of them are startups that are just entering the market and they try to limit themselves to using only, say, Landsat images that are free and the resolution is not so good and they cannot actually get really good results in this tracking of carbon offsets.

And this is a huge problem because the number of startups I’ve seen and the number of startups that became an advisor to, they try to save money. And the problem is that most of the satellite companies don't give some satellite images for free to these startups to try something. At least, it can be a small area. If they can have it for free, then they can develop their technology, become the client of the satellite company that supplies satellite images and that can bring a great boost to the market because nowadays they try to save money, as I said, and the results are not as good as they could be.

[Rachana, Host]
Besides being prohibitively expensive, is the current satellite imagery that is available, sufficient in terms of spectral bands or spatial resolution or the frequency of availability or even geographical areas of interest?

[Dr. Dmitry Mikhaylov]
Geographical areas are definitely a problem so far. Not everything is covered, at least in some countries where I worked. I worked in Uzbekistan, in an international program of tracking carbon offsets. They have a problem. They have these images available only once a week, say, in some of the areas. And it's a huge problem for the country because they have to prove that carbon offsets work.

Another problem is that, most of the carbon offset startups, they use different indexes for tracking vegetation like Normalized Difference Vegetation Index.. And these indexes are calculated wrongly because the quality is not enough, the frequency of the satellite images supplied is not enough. So they don't get very good data and based on this data, they cannot really calculate the proper formulas, they cannot calculate their carbon offsets. And the trust of the data is low.

[Rachana, Host]
Wow, that’s very interesting. In which other geographical areas do you think there is a serious lack or dearth of satellite imagery across the world?

[Dr. Dmitry Mikhaylov]
I worked in Central Asia, it’s a major problem there. Africa definitely as well. Latin America. All these three areas, where I have experience, they have a lack of access to good data. Moreover, it's quite important that, for example, working in Brazil and working in Uzbekistan, you have to implement very different approaches because they have so different climatic areas, and thereby different vegetation, different sunlight, everything is so different.

So you have to adapt the data you have, to the local reality. And the problem is that it can only be done by local startups who have historical data or have access to field measurements. They will have access to, for example, multispectral camera on drones. Because you need all these different layers of data, not only satellite, but drone images, IoT devices and soil measurements to make a proper model on calculating the carbon offset.

So it can be done only by local startup actually. But can these startups [function], if they don't have enough money. Say for example, in Uzbekistan, the investments in these local startups, in carbon offsets is almost zero. So these companies don't have enough budget, so they produce not very accurate calculations. And the country suffers because the country cannot really show that carbon offsetting is working. I would think about making a small proposal before the VC funds from satellite imagery producing companies on how to invest in small startups in these countries so that they can maybe provide them images for free so that they can really make proper calculations of carbon offset formulas in the region.

[Rachana, Host]
Why do you think there has not been interest in this direction so far? Or is it because the financial returns are not really expected to be that high, that lucrative for investors or satellite operators? Or is there any other reason?

[Dr. Dmitry Mikhaylov]
Yes, the payback is not so obvious. Regulations are evolving and changing all the time. Of course, it’s a very good area and most of the VC funds go there. But they mostly go to countries like the United States or Europe, because there it's quite obvious. Because a green agenda has been there for a long time.

And the Central Asian countries and other countries have other problems. And carbon emissions is perhaps not their top priority. And that will stay the same for a long time, unless we have other vehicles to sponsor local initiatives.

[Rachana, Host]
Yeah, absolutely. I guess political stability and clarity around space or satellite imagery regulations would also play a role.

[Dr. Dmitry Mikhaylov]
You know, offsets must meet requirements of permanence and the impact of projects should not be reversed in the future. And stability is a huge, huge criteria for this. And political, geopolitical instability is also making an impact on their carbon offsets.

[Rachana, Host]
We are currently talking about satellite imagery but taking a small detour. Do you think having affordable satellite connectivity would make things easier in terms of all that we talked about in these geographical areas?

[Dr. Dmitry Mikhaylov]
Absolutely, because this will lead to new startups coming into the scene, offering very specific knowledge. I'll give you an example. Uzbekistan where I worked, is a country of cotton and nobody so far studied the carbon credits of cotton because most of the country’s economy at least five years ago was built on cotton export. And the problem is, a lot of the startups that tried to make these measurements made quite impressive research in the area. And it will be an export product.

They could export these services to other countries like Pakistan, India, United States, China, everywhere, because they have immense data that they can use to train a very good AI model at least. But again, nobody invested in this case. This whole idea did not come through.

[Rachana, Host]
Wow, that’s very interesting. Being a satellite manufacturer, we never really think about these things or imagine these kinds of problems on ground. You said that even though there is a decent VC presence, there is not much investment into carbon credits across the world.

Do you think with the increasing awareness about climate change, do you think this whole scene is changing? Do you think it will push more resources to invest into supporting these carbon credit companies? And also, do you think the demand for real time satellite imagery will change? And how do you see all this panning out in the next 10 years, 20 years. I don't know. In the next ten years, 20 years.

[Dr. Dmitry Mikhaylov]
The demand will really increase for sure, multiple times. But the problem is that I think that VCs are not the power who will push this carbon offset market because VCs follow regulations and follow market demand. And the market in this case is formed by regulations in the country and between the countries. And we have to focus on these stable rules, clear rules on regulations. Then the VCs will come and then we will have a huge demand.

There will be a lot of startups and there will be a huge demand for the satellite data. And not only startups, it can be government government companies as well. But regulations come first, same as in cybersecurity, by the way.

[Rachana, Host]
That’s very interesting because historically when we look at the evolution of the whole space era for humanity, it was always regulation that was following space technology. Now it looks like regulation has to take the lead in propelling us forward.

[Dr. Dmitry Mikhaylov]
Absolutely. It's my strong belief that only regulations will make it happen.

[Rachana, Host]
That's a lot of very insightful, very useful inputs you have given Dr. Mikhaylov. Like I said, someone who is in the echo chamber of the space bubble, And especially as engineers, we are just too occupied with the engineering aspects of building satellites.

But it’s very great to know all the repercussions, how everything flows, the ripple effect of regulation and how it can impact technology. That’s very interesting and insightful. And thank you very much for your wonderful inputs, and thank you very much for your time.

[Dr. Dmitry Mikhaylov]
It was great to have you on the podcast.
Thank you for inviting me. Thank you.