Rad-hard: one company is keeping space systems operating years beyond expected lifetimes

In this Q&A with Steve “Doogie” Russell, who leads Space Systems for BAE Systems’ Electronic Systems sector, we discuss BAE Systems’ history of radiation hardening from Apollo to Mars rovers, some of the science behind rad-hard and rad-tolerant systems, and the new role radiation hardening will play in enabling small satellites in proliferated low-Earth orbit (LEO).

Breaking Defense: Tell us about BAE Systems’ space history and contributions in the area of radiation hardening that go back to the Apollo days.

Russell: Our heritage does go back to the earliest days of the space program. It spans everything from Apollo through multiple generations of planetary space exploration programs, as well as commercial and Department of Defense satellites that provide critical services to hundreds of millions of people today. Think GPS, communications, weather data, and even satellite radio.

Going back to the 1960s, BAE Systems, which was known as Sanders Associates then, built the Saturn V pre-launch checkout system, which was the first automatic computerized system in the space program. Apollo 11 incorporated BAE Systems’ displays, consoles, data processing capabilities, and even some electronic filtering capabilities that were used in subsystems on the lunar excursion module.

Since then we’ve flown on a variety of NASA missions that include the Cassini space probe to Saturn. Four different generations of rovers on Mars, several of which are still operating today — for example, the latest Perseverance rover — include BAE Systems’ products. Most recently, the James Webb Space Telescope relies on BAE Systems’ computers and Application Specific Integrated Circuits for instrument operations and spacecraft command and data handling work.

Breaking Defense: Are there quals you can point to like Service Level Agreements out to six nines, for example, that demonstrate how BAE Systems’ radiation hardening has helped space systems fulfill their missions over many years?

Russell: Across this spectrum of capabilities we’re discussing, we’ve never had a single known failure and we’re up to about 11,500 years of operating time on our systems that are in space.

Take Cassini. Once it was launched to Saturn, there was no servicing the probe if anything went wrong. Its critical functions had to work reliably all the time because if it fails you’re done and your billion dollar mission has failed.

It’s the same for the latest Mars rovers, Curiosity and Perseverance. Curiosity landed on Mars in 2012 and was designed to have a two-year design life. More than 10 years later, it’s still operating. Perseverance has been operating on Mars since 2020 with BAE Systems’ RAD750® single-board computer from the same basic product family that was on Curiosity, and has been the “go to” solution in high-reliability radiation hardened computing in space for the last 20 years.

Breaking Defense: We know that systems in space have always needed to be hardened against radiation. How is hardening typically done?

Russell: I’ll lean into the specific answers here in just a second, but I think it’s important to understand what’s causing the need for radiation hardening when you’re in space environments. To your point, you can’t just take a normal processor like you have in your laptop and put it on orbit or in deep space and expect it to work without being disrupted.

Radiation-related disruptions or upsets are caused by a variety of different, highly energetic particles in space that are radiating from various different sources. Those particles actually go right through the outside parts of your satellite buses. They will free up electrons and other energetic particles that can disrupt the operation of the electronics, penetrate the processing and memory systems, and even the silicon itself.

There are a number of different degrees of upsets that can happen. Some are temporary that can be fixed with a reset. Some of those disruptions, however, end up being fatal disruptions that we call latch-ups where it would not just shut down the circuit, but shut down where you could never start it again.

This is more critical in higher orbits and in deep space than it is in lower earth orbits, and we’ll talk later about rad-tolerant capabilities that are specifically tailored for those lower orbits.

With that overview, there are three major steps associated with taking something that’s not rad-hard and making it rad-hard. The first one is the requirement specification, and that is understanding the function that the processor or other electronic component has to perform, as well as the environment such as deep space or low-Earth orbit.

A subdivision of that is the criticality of the function. Is it a function that absolutely cannot ever fail otherwise the mission fails? Or is it something that could be a temporary failure that could be fixed with a reset at a system level?

The second step is actually designing your radiation hardening or mitigation. There’s a variety of different things that you can do. It could be a physical design change that spaces out transistors or other electronic components on a silicon chip so that a single radiating particle won’t disrupt all of them at the same time. It might be a logical design where you have redundancy built into your processing that wouldn’t normally be there or required inside earth’s atmosphere.

There are also system-level mitigations versus component mitigations. An example would be error correction codes. If you had an upset in one memory slot, can you fix that if you have appropriate error correction codes designed into the component or system?

The final part of that second design step is designing packaging and shielding for board-level systems up to the larger assembly level, as well.

Once we’ve done all of that, then you build it and put it through verification and validation testing. We have radiation sources that we use to test the lifetime of the system to make sure that it performs the function expected.

Breaking Defense: What’s new on this front? What are the new technologies and capabilities that improve rad hardening?

Russell: There are two major trends in the industry that are particularly exciting and driving the rad-hard market. First, there’s high demand right now for significantly more processing capability at the same levels of radiation hardening that we’ve had in the past.

I mentioned earlier that BAE Systems’ rad-hard solution for the last two decades has been the RAD750® single board computer. We have a new version out now that’s called the RAD5545® single board computer, which is the most advanced and capable radiation-hardened space processor in existence today. The microprocessor integrated circuits at the heart of the computers are based on different generations of integrated circuit fabrication technology, what the integrated circuit folks call nodes, which refers to the minimum feature sizes of individual components on the chips.

The RAD750® computer is based on the 150-nanometer node. The RAD5545 computer is based on the newer and more advanced 45-nanometer node, meaning that the minimum size of its features is 45 billions of a meter. The smaller, lower power transistors of the 45-nanometer node support higher function and performance. What’s exciting right now is getting to higher processing throughout in the same size and with the same level of reliability but in much smaller geometry nodes. We’re getting closer to catching up to what they do in terrestrial computing domains, where they have even smaller nodes.

The next major node is the 12-nanometer node. In 2021, BAE Systems announced that the Department of Defense has contracted with us to develop radiation-hardened libraries in this 12-nanometer node. Those are going to enable a significantly higher amount of processing power on orbit at the same level of reliability that we’ve had in the past.

The second trend is almost the exact opposite of the first trend, which is that there’s a shift within the market to complement legacy space capabilities and big systems presently in orbit with small satellites in proliferated constellations in low-Earth orbits. It frees up a huge number of different design trades for the folks that are building those constellations.

LEO satellites tend to be smaller and they don’t need the same level of radiation hardening that you need in a GEO orbit or deep space. Operators are targeting lower costs because if you want 100 small satellites then you want your processors to be cheaper.

Also, because they don’t have to be as radiation-hardened in LEO, you can use more commercial-off-the-shelf types of processing capabilities and electronic components and then do a lesser amount of shielding or other radiation hardening techniques — certainly not at the same level you would do for high-end radiation-hardened capabilities.

That allows us to take the technology we’ve developed for radiation hardening at the extreme level and then apply it to up-screening commercial parts that are more readily available and have more capability. From the operator side, that frees up our customers to do things that they couldn’t do with the traditional high-end radiation-hardened types of processing capabilities.

Breaking Defense: The trend toward small sats in proliferated LEO will drive development of radiation tolerant solutions. Explain.

Russell: This is an exciting and rapidly growing part of the space electronics market because of the capabilities that we can deliver from P-LEO types of constellations. Unlike rad-hard solutions, a lot of these are designed to be lower cost and have lower reliability.

That sounds weird, but these systems are not required to operate for 10-15 years with zero upsets. They may be designed to operate for only three to five years, and because they’re lower-cost systems, you can get away with having a lower level of radiation hardening because they’re designed to be replaced.

As I mentioned, this is opening up the design trade space to have lower cost points and more capable processing that lets you conduct different missions from LEO constellations than those you can do from a large, single satellite in a different orbit.

For example, BAE Systems is known as a capability provider in the air, land, and maritime domains as well. What this allows us to do now is take the operating capabilities that we have in those domains and push them up into proliferated LEO orbits to provide more capability for our customers.


Leave a Reply

Your email address will not be published. Required fields are marked *