WASHINGTON — Lockheed Martin is intricately tied to the Pentagon’s future space endeavors. The giant defense contractor has deals for the Air Force’s next-generation missile warning satellites, it’s new batch of GPS satellites and the current generation of protected communication space vehicles.

But the national security space community is changing fast. Space is now viewed as a war-fighting domain, a far cry from decades ago.

Rick Ambrose heads the company’s space division. He spoke with Mike Gruss, editor of Defense News sister publications C4ISRNET and Fifth Domain, about where the Pentagon is headed and how to make sense of the new realities in space.

What advantages do you see with the Air Force’s new missile warning satellite program over the current Space Based Infrared Program?

The Next-Generation Overhead Persistent Infrared Program is, in essence, a whole new design, which is why we’re competing for payload. So it’s going to have a tremendous new mission capability, built-in resiliency capabilities, much more flexibility.

What does that mean, flexibility?

We’re going to put in some of our smart sat processing that will help with the payload. We’ll make sure we can incrementally upgrade, or the Air Force can, over time.

If you think about this, SBIRS [the Space Based Infrared Surveillance system] was originally designed back in the ’90s. Basically it’s a whole modernization of the mission — better performance across the board.

We need more continuous coverage; you need better resolutions. You need a better differentiation of the threat. You need to build in the resiliency, plus the modern ability and some of the processing. So how do we upgrade algorithms on the fly? All that’s going to be enabled in this design.

When we talk about the smart sat part of that, is that something that today you would get an image and then have to process it on the ground? So the savings is you can do it there so you get it faster. Or is there a different advantage?

This is always the trade-off. To process everything on the ground, you have to now communicate every piece of data down right away. We still may ultimately want to do that. But what if we can run some processing on the satellite versus the ground? That design’s still not perfectly baked in yet, but that’s the direction we’re going, is to build in some of that.

I think of it like adding filters on Instagram.

Another way to look at it would be: There are certain things that you’ll locally process on your phone not to clog the communications.

We can upload patches and software like we do on most satellites. We’ve been doing this for decades. But now it just gives you more flexibility to do even more things.

You know, a lot of times we’re flying satellites for 20 years and we keep finding new ways to use them. Let’s build that in up front.

I would imagine the Air Force is more open to that kind of thinking.

Oh, absolutely. Well, because the threat environment has changed, there are go-fast initiatives, [such as Space and Missile Systems Center] SMC 2.0. We love it because things in the past, it would take longer to prosecute changes on.

Now with their new push — you know, [the assistant secretary of the Air Force for acquisition, technology and logistics], Dr. [Will] Roper’s push for speed and agility — now once we get the program going and get all the designs finalized, then you’ll go to a more traditional —you have to prove out the concept and prove out that the system works and then deploy.

We do agile develop for them, but they go: “We’re going to constantly change.” Well, commercial could get away with that because if suddenly Google goes down, you don’t lose lives. These systems protect lives. The men and women that serve, weather systems, even the private citizens. It’s serious business. So that’s where we’ll always be different than some on the commercial side.

If you think about timing of the program up front, at the stage we’re at [a low-production rate]. Now is where we can do experimentation and try out new designs.

With the Space Development Agency, how do you see that integration improving?

What happens is the exact opposite of what should happen. Let’s say it takes five or six or seven, eight years to get a satellite up. That’s an expensive item. We have to move out and let’s get the satellite going. Well, nothing ever works that simple. What we’re saying is you need to put the end-to-end architecture together.

That’s why we brought our ground system together with us to help us help the government with this challenge. And then you get faster. And the other side of this, because it’s on the ground, you go: “Oh, it’s on the ground. We can always fix it.”

The hope is with the Space Force, [SMC 2.0] and all this, we can synchronize better. But more importantly, how do we make a lot of technology more common to the space and ground infrastructure?

If you’re having to develop every element of that from scratch, it’s just massive, it’s costly.

So what can you do?

I did a study decades ago because everyone concentrates on the satellite. I said: “Well, what’s the ground cost?” I ran our satellites and we’ve designed them to run 20 years. You go: “OK, what’s the infrastructure cost around that?”

And when you took a 20-year cycle of the ground and operations and processing, and think about it, every three years or so they’re upgrading. Because you have people touch your computers so now they got to upgrade the machines every three or four years. The IT infrastructure and all that. Refurbish all that.

The cost of that dwarfed any costs over that time period of the space asset because you paid once.

It actually was more expensive than all the satellites and launches combined. We can knock the ground back a little bit by putting artificial intelligence in, ultimately machine learning, more automation, simplifying operations.

You mentioned resilience at the satellite level. There’s been talk: “Could a satellite evade a missile? Or evade another satellite?” People have a difficult time understanding what resilience at the space level means.

If you’re thinking of resilience, it’s going to come in a couple of flavors. You touched on the first one.

First, if you set your architecture up right, it’ll inherently give you some resilience and allow you to make some different trades on the satellite level. Then the satellite itself can just be much more robust.

So just inherently for mission assurance, the satellites are more robust and we’ve put redundant systems at higher quality, higher-reliability parts. You can think of it that way.

For resiliencies, you well know there’s some level of hardening on SBIRS and the Advanced Extremely High Frequency satellite anyway for both environmental as well as man-made events. The best it can do.

Think safety systems you’ve put on your cars. In the past, when there were a few cars on the road, no one really worried about it. I think the first cars didn’t even have seat belts. Or you just keep adding features as you learn more things. It’s like with cyber, everyone says: “When are we done with cyber?” You’re not because it’s a journey. Every time you do something, someone else tries to defeat it.

Boeing is under contract for the Air Force’s next wideband communications satellite. The company is trying to quickly build it. Are you watching that process?

We can come up with our ideas, some other people have ideas.

The thing that is just fantastic about space right now [is] it’s no longer just competition of its traditional players. There are over a thousand new entrants now if you count the numbers. You got large players coming in like Jeff Bezos. You’ve got traditional competitors, you’ve got the supply chain forward integrating. Think of a Harris and L3 combination.

Those are all competitive surfaces, which makes this industry just damn exciting right now. And it may sound crazy, but that stimulates motivation. It stimulates innovation. It simulates the thinking and those competitive spirits, where it’s kind of what this country was founded on, right? So we’re always watching that.

We’ve really modernized our production. In the old days everyone would hand-lay down the solar cells. We now have robotics and automation equipment just literally laying those cells down. It’s more predictable. It’s more ergonomically friendly for my technicians. One cell was like art almost. Now we’re trying to say: “OK, we don’t want to lose performance, but let’s build in the design for producibility, operability, operations [on] Day One so that we can automate it.”

So let’s say an electronic card, which would take a technician three months to put together, solder, fill and now we run it down the line; in under eight hours, it’s done.

Is the Air Force OK with that process? I think of this as pretty unforgiving.

Well, it still is. You go back six years ago, I think we did a dozen [3D]-printed parts. We did over 14,000 last year.

If you go through our space electronic center, we put automation in. The problem for us in space is we have some unique parts and they weren’t precise enough. How do you measure it? You know it’s very valuable. You know you’re taking time out. If anything, you’re improving the quality of work life for your employees.

There’s this discussion that GPS III is the most resilient GPS satellite ever. And at the same time, the Army says: “We should count on it less than we ever have before.” How can both things be true?

In GPS III, it’s a much higher power. The M-code coming online makes it somewhat more resilient. But you’ll still — again, just like cyber — you’ll have adversaries still trying to figure out engineering and different things, techniques.

If you take your GPS commercial receiver and you’re running in the city, you get a lot of bounce off that urban canyon. So it knows like: “What? That dude looks funky. Throw him out.” Then it processes the ones that it thinks are good. That’s a form of protecting that environment if you think about it.

How will this play out?

There’s going to be some combination of software and then maybe some other sensors like we’ve been toying with, some microgravity sensors, which you can then kind of tell the region you’re at.

And some of the — just the onboard inertial systems — are getting pretty damn good. It’s like your self-driving cars. It’s going to rely on not just the cameras, but the little radar sensors and some combination of sensors.

For [timing], when you’re running software and you have all these sensors that are nodes in the network, and they can actually talk to each other, this is maybe a nirvana future state. Then the guesses you make are better informed with more data.

There could be a world where GPS is making decisions with 80 percent of data that’s coming from GPS satellites, and maybe it’s pulling something from some other sources.