BlackSky is a geospatial intelligence company taking advantage of the economics of small satellites to efficiently carry out global monitoring. But according to BlackSky CEO Brian O’Toole, the company “is more of a software data analytics company that happens to own some satellites.”

In December of 2020, Constellations Podcast sat down with Brian O’Toole to discuss the interesting things BlackSky is doing with software data analytics. In this exclusive interview, O’Toole sat down with Constellations again to dive deeper into the concepts of cloud computing, Artificial Intelligence (AI) and Machine Learning (ML), as well as the company’s plans for provisioning and launching more satellites to support its Earth Observation (EO) operations.

Constellations: For those who haven’t listened to the Constellations podcast episode – in a nutshell, what does BlackSky do?

Brian O'Toole: BlackSky is building a high revisit, high resolution small satellite constellation. The goal is to enable near real-time observation of the Earth. We've combined that with a software platform that can deliver real-time intelligence imagery and data analytics around the world.

In a nutshell, BlackSky is more of a software data analytics company that happens to own some satellites. Obviously, satellites deliver space data on demand, giving us the ability to observe global events as they're unfolding on the ground ... but it's the software platform that gives us scalability and opportunities for growth, and that's driven by where the market is going. Its customers are ultimately looking for real-time geospatial intelligence, not just imagery … Even though satellites are an important part, the software is a key part of our strategy. The platform enables us to deliver data from many different sources and apply artificial intelligence and machine learning algorithms to derive insights that matter.

Constellations: What has changed for BlackSky since constellations last spoke to you in December?

O'Toole: It has been an exciting few months. We announced our SPAC transaction a few months ago – it fully funds the company and puts us on a path to realize our growth.

We also just announced the successful launch and commissioning of our latest satellite. That all occurred [recently] on a Rocket Lab mission – then around the same time we announced an agreement with Rocket Lab for our next five missions, enabling us to accelerate our satellite constellation deployment during 2021. So, an exciting first quarter.

Constellations: Speaking of BlackSky signing with Rocket Lab again, what added value will the satellites bring the current constellation?

O'Toole: The Rocket Lab announcement puts us on a clear path to launch at least another 8 satellites this year, which would be added to the 6 that are operational today. That puts us on the path to a 14-satellite constellation by the end of the year. This gets us much closer to where we want to be, which is the ability to achieve an average of one hour revisit, with the ability to provide dusk to dawn imaging capability, as well as deliver the information and analytics in around 90 minutes or less.

Constellations: What are your plans for launching satellites in the future? What new technologies will be used?

O'Toole: As I just outlined, we plan to expand our Gen 2 constellation, which is our current satellites with meter imaging capability. We also announced an investment in our Gen 3 satellites within the last year. Those satellites are already in development and moving forward, and will start coming into our constellation in 2023. That’ll bring our imaging resolution down to 50 centimeters ... the expansion of those satellites puts us on a path to a 30-satellite constellation.

Constellations: The podcast episode touched on AI and ML, and how these technologies are being used to provide insights to customers. What current capabilities do you have now, and how do you see them developing in the future?

O'Toole: We do a lot of machine learning on imagery … Looking at that data over time, it gives us a historical view of what's going on in a particular location. This moves us into the next dimension of understanding what's normal and abnormal behavior for a particular location – for example, how many trucks are typically crossing at this border location on a given date or time.

Machine learning and AI elements apply to everything from the information we're extracting out of the imagery, to the time-based insights and analytics – we're getting some understanding of what's going on ... We also do quite a bit using natural language processing. We're looking at new social media and other data feeds of things that are being broadcasted, or data streams from around the world, and we're analyzing those in real time in different languages. We're using that to detect and find activities as they're unfolding. As we're able to find those things, we can automatically tip and cue our satellites to collect information about the event as it's happening on the ground. So there are lots of different applications for AI and machine learning, not just in the imagery, but in all different kinds of other data.

Constellations: Another concept that was mentioned in the episode was cloud computing. What advantages does cloud computing bring to the geospatial intelligence arena?

O'Toole: In general, the advent of AWS [Amazon Web Services] gives us access to near infinite computing resources. We are not hindered in the scaling of our computing infrastructure, to not only deal with the scale of data that's coming from our constellation, but also the algorithms that we're applying to that data to discern insights.

Then finally, through that same infrastructure, we can deliver that to customers on demand around the world. There are a lot of advantages, and it accelerates our speed to market and our speed to insights ... we're not worrying about scaling or computing infrastructure – it's available for us as we need it, and as the business grows.

Constellations: Related to cloud computing is the concept of virtualization. With the move to software-defined satellites and the virtualization of the ground system, what advantages and challenges does the geospatial intelligence segment face?

O'Toole: Good question. Software virtualization helps from a variety of different dimensions. It means you can package up your software and deploy it differently for different customers. It also enables you to scale.

In our case, our platforms are designed to run in a Machine to Machine [M2M], fully autonomous type of configuration. When you get into virtualization and some other technologies, it’s just a new tool to allow us to scale our software-based capability. The satellites themselves are more and more software-defined, which gives us a lot of flexibility operationally. From maintenance and upgrade, to our ground network in terms of our receiving stations, and then everything we're running in terms of image processing, delivery is all driven by a highly scalable software environment. Cloud computing, combined with virtualization, is a powerful enabler.

Constellations: How do you see the geospatial intelligence segment developing over the next decade, from both a technology and business standpoint?

O'Toole: From a tech perspective, there's going to be more sensors. Not just satellites, but IoT [Internet of Things] sensors – and there's an ongoing growth in drones, etc. More sensors will give us more raw data for geospatial intelligence. Then, the analytics and software platforms being developed to process that [data] and turn it into meaningful insights are just going to get better, providing deeper information faster and faster.

I think the geospatial intelligence industry is entering into its next phase … in the last five years or so [the industry] has moved more into location-based services, and now I see it moving more into real time intelligence that's accessible to a much broader commercial market. That's just going to enable a whole range of new applications and adoption by many new types of businesses that traditionally haven't used this type of data before.

On the government side, another big shift we're seeing in the market is the move to tactical ISR [Intelligence, Surveillance, and Reconnaissance] from space ... I think satellite technology in the past has been primarily used for intelligence applications. Now, we're getting closer to real-time high revisit collection capability and rapid delivery that can start to be deployed to support the warfighter, and more tactical applications.

Constellations: In wrapping up the interview, is there anything else that you'd like to add?

O'Toole: BlackSky is at an exciting time right now ... and I think it's an exciting time for the industry. Space is still a high barrier to entry market. It [takes] a lot of time, capital, and critical talent ... but I think that with the recent investments in everything from launch, to communication networks, to satellites, to analytics, we're seeing industry moving into its next phase beyond NewSpace.