Military Embedded Systems

Small satellites increasingly tapping COTS components


June 08, 2015

Sally Cole

Senior Editor

Military Embedded Systems

Many of the COTS technologies prevalent in today's low-cost ground applications are now being adapted for space and making their way into small satellites.

Improved launch access to space is helping to drive the small satellite market and is behind a new focus on using small systems. Now that it’s possible to readily place small satellites on launch systems being sent up to the International Space Station or to low-Earth orbit (LEO) at a fraction of the cost compared to in the past – for hundreds of thousands of dollars now, rather than millions – and on a fairly regular basis, people are finding unique ways to use space and create new disruptive capabilities.

Small satellites are broadly defined as those having a mass of as much as 500 kg. “There’s a continuum of intertwined sub-demarcation that corresponds to increasing mass and volume, including nanosatellites, which are typically referred to as ‘CubeSats.’ CubeSats range from one to 10 kilograms and are described in canonical units of 10 by 10 by 10 centimeters,” explains Aaron Q. Rogers, a small-satellite expert at the Johns Hopkins University Applied Physics Lab.

COTS in space

Interestingly, commercial-off-the-shelf (COTS) technologies are playing an increasingly significant role in small satellites. “Clearly, as we see a move toward commercial endeavors with fairly aggressive cost points for small satellite designs, it involves using industrial- and automotive-grade electronics and other elements from nontraditional space markets … so COTS use is becoming much more prevalent,” Rogers says.

With COTS, no parts can be used whimsically – every part still requires pragmatic selection and screening. “And you need to reconcile that process with your expectations for the success and outcomes for those parts,” Rogers points out. “We’re seeing a lot of electronics – imaging technologies, radio technologies, navigation and GPS receivers, and other things we take for granted in our cellphones – moving into space designs.”

The key to using COTS parts is to not need to modify them to work in your system with any repackaging. “You want to take advantage of their cost, so it’s sort of intrinsic to buying them off the shelf. I can speak to experience that we’ve taken COTS components and had to shrink them in one dimension and it can become very expensive,” Rogers adds.

Testing can vary

Sometimes small satellites require very little testing, “driven by both a high risk posture in which the design can be made so cheaply so that there is minimal financial impact of a loss, or it’s schedule-driven to support market opportunities so testing needs to be made simple and efficient,” Rogers says.

There are, however, other cases in which greater priority “is placed on ensuring the satellite will work for an extended period of time with a higher confidence level that it will deliver its designed capability as expected,” points out Rogers. “So even if there’s a decision to use commercial parts, these mission developers may put them through additional screening levels – derate them, meaning operate them at levels lower than they’re actually specified to provide some headroom on how much they’re exercised, or using redundancy and things of that nature. It’s a bit of a continuum that depends on the application, what the customer or the sponsor demands, and, of course, the cost and schedule.”

Historically, a detector or optical system needed to be run through a series of tests or had to be of a certain quality or greater. “Now, some companies are taking systems more akin to those used in automotive or industrial applications – with a pedigree for operation within severe environments with thermal extremes and shock – and finding that they can translate them to space applications,” Rogers says.

There’s still plenty of physics and work that goes into, for example, ensuring parts such as cards are radiation-hardened for space. Embedded computing company Aitech Defense Systems Inc.’s next-generation microprocessor, the SP0, uses Freescale’s MPC8548e processor, which they “worked on with NASA and used the cyclotrons at the University of California-Davis to subject all parts and products to radiation testing,” according to Doug Patterson, vice president of the Military & Aerospace Business Sector for Aitech (Figure 1).


Figure 1: Aitech’s SP0 3U CompactPCI single board computer passed 100 kRad (si) testing.

(Click graphic to zoom by 1.9x)




New ways are being discovered to adapt COTS technologies that are prevalent in low-cost ground applications for use in space, Rogers says. “It’s really more of an acceptance of risk and finding ways to maneuver around and mitigate that risk to be more successful and to realize all the cost and scheduling savings that can come from it,” he adds.

Another trend within this area is “a new move afoot within NASA and at the big primes to use non-EEE (electrical, electronic, and electromechanical) parts,” Patterson says. “While it hasn’t gained much traction yet, it’s significant because these are the most reliable parts available today. The tradeoff is that EEE parts are extremely expensive, so reduced cost enables you to build more satellites and possibly put them into LEO and interconnect them through a wireless network capable of loadsharing.” (Figure 2.)


Figure 2: Although three cans of soda would fill this Firefly CubeSat to its brim, NASA has big plans for these tiny satellites. Photo courtesy of NASA/Bill Hrybyk.

(Click graphic to zoom)




Redundancy concerns

Do you need to worry about redundancy when using COTS parts? “Yes, maybe. It really depends on what you’re trying to do or what your approach is,” Rogers explains. “If you have a mission that needs to work for one to three years or more, or if you’re using a lot of parts that have less traceability in terms of parts’ pedigree, performance, or reliability, one way to overcome the uncertainty is to fly redundant parts and have spares onboard so that if a part fails you can bring the backup online. This isn’t typically done, but flying extra copies of a memory unit or power system or something is a way to work with commercial parts.”

While there are many commercial forces pushing the cost model toward very cheap providers and suppliers, “some missions still require high assurances of performance success – particularly, ones that the military or government count upon operating as expected,” Rogers says. “For those, there’s still a need to do good testing and to use good parts and have a well-thought out, pragmatic, and enduring design approach.”

That’s not to suggest that these two types of missions can’t coexist and hopefully benefit from each other, “because the products and offerings that fit the needs of a more high-performance low-risk mission can also move downward to support the more cost-driven missions as production scales and qualified low-cost COTS solutions can migrate upward,” Rogers notes.

Right now, a bit of bifurcation exists in which some players entering the small sats space can’t afford to lose a mission for political reasons, unlike commercial endeavors that can go up and take gambles and learn on the fly, then iterate. Rogers says he expects to see these two communities “probably further converge within the next five years. It’s going to be a real change in terms of how space is used and accessed,” he adds.

A number of small companies – such as Rocket Lab and Firefly Space Systems – are trying to build dedicated launch systems for small satellites and are receiving big investments from venture capital firms. “When these come to market, and they have credible goals of flight demonstrations within the late 2015 to 2016 time frame, then we can really start talking about where we want to go for these small Earth missions to create constellations of purposely built deployments and really bring the cost bar down even further,” Rogers says.


Sidebar 2

(Click graphic to zoom)




Rightsizing small sats

CubeSats are very small – nominally four to five kg for a 3U or “triple” – and limited in volume, but “it turns out that making them a bit larger by moving to a 6U form factor or as much as a 50 kg ’Express class‘ small satellite, the additional cost for manifest doesn’t typically directly scale as long as you can still take advantage of launch access and rideshare missions,” he says.

This means that greater volume “enables greater use of COTS components, which provides more performance or accommodation to sensors and payloads. This could be a tremendous value position that we, as a community, are figuring out right now,” Rogers notes.

It also appears an inflection point has been reached in terms of small satellite size. “A good analogy is comparing the right size for small sats to iPhones or other smartphones,” Rogers notes. “They kept getting smaller until they became too small, and now we’re moving back to slightly larger sizes to gain a little more capability, performance, and utility.” (Figure 3.)


Figure 3: A set of NanoRacks’ CubeSats as photographed by an Expedition 38 crew member aboard the International Space Station. The CubeSats program contains a variety of experiments, including Earth observations and advanced electronics testing. Photo courtesy of NASA.

(Click graphic to zoom by 1.9x)




Small satellites also bring new security concerns. Satellite command and control (C2) links require encryption or other data privacy methods to prevent intrusions and unwanted access. “Keeping control of your C2 networks is very important and people are actively working on methods that are effective and consistent with the more limited resources of small satellites,” Rogers notes.

Military applications for small sats

Plenty of missions are now driven more by high temporal or spatial access rather than by exquisite performance from a large platform with a single aperture or sensor. “This means it may be possible to support a mission with a very modest sensor – a radio or a camera system – by using lots of them,” Rogers says. Creating constellations of these systems enables huge revisit rates or access to areas on the Earth in a way that can provide a lot of details at a heightened repeat cadence. “We’re seeing this on the commercial side with global imagery and weather … and it’s sort of driving itself backward into the Department of Defense side,” Rogers continues. “The government is really starting to take note of opportunities to supplement their systems with these additional data sets to improve overall knowledge. Constellations are currently being explored for automatic identification system (AIS) tracking applications, for example.”

For its part, the U.S. Air Force Space Command (AFSC) is exploring several cost-saving measures like smaller satellites, disaggregation, commercial leasing arrangements, and block buys, which provide the benefit of economies of scale.

Functional disaggregation “means that sensors or submissions previously contained on a single satellite are now dispersed across several smaller, less complex, and more affordable satellites,” explains Anthony D. Roake, chief of Current Operations Division, AFSC Public Affairs. “AFSC is interested in the concept of disaggregation because it can increase resilience by dispersing capability across smaller, less complex satellites to avoid single-point failures that could cause a catastrophic outage over the battlefield.”

The Air Force’s Operationally Responsive Space (ORS) Office is interested in capabilities to get on orbit faster and cheaper, and some of its programs involve CubeSats. The AFSC is now “taking advantage of lessons learned from the ORS program to explore building smaller satellites, some of which represent disaggregated capabilities and others that would merely take advantage of smaller sensor packages,” Roake adds.

Increasing use of COTS and open standards ahead

For the sub-100 kg regime of CubeSats and small microsats, the community of suppliers for space-qualified components is currently limited.

“Many developers are trying to build their own or use COTS parts in new ways, and as those are realized to be successful – or lessons learned are brought back and iterated on to make the design more suitable for space applications – this market should stabilize and grow,” Rogers says. “Similarly, because there are limited vendors to draw upon as a satellite developer, not many standards exist at this point, so people doing design work tend to do whatever makes sense in their individual case. There is some consistency on electrical and data, and, in some cases, mechanical interfaces, but there’s still a lot of variance involved.”

Commercial companies are “building CubeSats now, so we’re starting to see them give the large primes some competition,” Patterson notes. “But there are plenty of opportunities – hardware is still expensive and no one wants to send it to space only to have it drop out of orbit.”

Moreover, as Patterson points out: “Companies with experience producing space products – ones with commercial as well as military and space experience – are able to offer customers the choice and ability to move back and forth between components or to even selectively choose a component that seems to be less susceptible to radiation than others.”

Right now, “we’re in the nascent stage where many people are entering the mission development market,” Rogers adds. “As subsystem suppliers start to converge and drive consistency across standards interfaces to make their product offerings more producible and interchangeable, some of those things will converge, similar to what exists in the larger-scale aerospace industry. Expect to see this stabilize within the next few years.”


Sidebar 1: The Vermont Lunar CubeSat - launched more than 18 months ago by a team at Vermont Technical College together with NASA - is now only the one of the original 12 university CubeSats still in orbit.

(Click graphic to zoom)