Blog

Podcast: Is COTS now acceptable for space applications?

April 06, 2020 | BY: Jerry Festa

While the use of readily available commercial electronics, or so called “commercial off-the-shelf” (COTS) products, are commonplace in military applications, the same is not the case for space applications. Though this appears to be changing, there is still a concern about COTS reliability in the space community. Back in 1994, a famous memo was issued by William Perry, who was at the time the U.S. Secretary of Defense, which became known as the ‘Perry memo’.  The memo directed the department of defense to use COTS products where possible rather than custom built solutions. This was issued over concerns that using detailed military specifications was limiting competition, stifling innovation and increasing costs and time to deploy new systems.

So why aren’t COTS products now commonplace in space applications? Well, it is certainly becoming more so, but some think there is an anti-COTS feeling that has been entrenched in the more experienced members of the industry who are more comfortable to using space grade equipment. Historically, there have been problems with the reliability of COTS systems, and notably, there were projects that went to Mars in the late 1990’s that used COTS hardware that failed. However, although it was later discovered these specifically weren’t due to problems with the COTS hardware, there have been other failures which have sown seeds of doubt. Additionally, to make sure a system as a whole is reliable, space COTS systems may require additional redundancy and testing, eroding cost savings.

In general, the use of COTS has been somewhat cyclical. It gets used, it works well, then a problem happens and it gets used less. The industry then rectifies the problem and off you go again. It can be argued that at our current point in time, the parts and the processes being used by leading space COTS providers are now well proven to perform without failure and are being trusted in an increasing number of missions.

Another significant motivator for the recent boost in adoption is the rise in use of smaller satellites (small sats) in low earth orbit (LEO), as opposed to larger satellites in geosynchronous earth orbit (GEO). Small sats are compelling for a number of reasons. They are cheaper and easier to get into orbit, so instead of spending a billion for one GEO sat, you can put a constellation of small sats around the globe into LEO for same cost. They are also far cheaper to replace. It also makes sense to move from a 15 year technology refresh cycle that you would have for a GEO sat to a five year one (typical for terrestrial aircraft). This helps to ensure you are getting to access the capabilities of more modern technology. Therefore, since deploying to LEO means lower cost, it makes much more sense to use lower cost systems where you can.

All this is leading commercial, and an increasing number of government, organizations to using space COTS systems to remain competitive. A more detailed discussion on these, and many more issues surrounding COTS hardware and software, and the history of its use in space applications can be heard on the “Enabling COTS in Space” podcast, available from The New Space Race podcast series hosted by Wind River®.

Listen to Enabling COTS in Space - podcast by Wind River

 

Transcript

John McHale
Hello and welcome. My name is John McHale, Group Editorial Director for Military Embedded Systems Magazine and you're listening to the New Space Race - an original podcast series about the people and technology that are transforming the space industry today. The podcast series is brought to you by Wind River and for more information and to listen to previous broadcasts, just visit windriver.com/new-space-race/. That's windriver.com/new-space-race.

And now welcome to episode 3 - COTS in Space. The procurement term COTS - also better known to many as commercial off-the-shelf, brief little nugget here - it's been around since 1994 when then Secretary of Defense William Perry instituted a memo saying we will buy COTS wherever and whenever possible and because of a lot of the bad press they were getting over 400-dollar hammers and thousand-dollar toilet seats.

Well now 25 years later COTS is just the way many people do business anymore but not so much in the space industry. Always, like everything else, it takes a little longer for us to get stuff adopted in the space and unfortunately the word COTS is a four-letter word in among the space folks. They consider it taboo mostly because of the commercial part of the phrase as they feel it connotes low-quality or lower viability - ironically saying things many people in the DoD thought 25 years ago. How times change. But today many COTS solutions are not low-quality at all and we're seeing them in demand for mission-critical space applications.

So our guests today - Jerry Festa who is the Senior Product Line Manager for the space segment, Curtiss-Wright Defense Solutions and Mike Deliman, who is the Principal Technologist with Wind River, will discuss how the challenges brought by this increased use of COTS hardware and software, how we can overcome them and how the increased use is basically changing the way people do business in the satellite and spacecraft industry.

So before we get started I would like each guest to kind of talk a little bit about their roles at Wind River and Curtiss-Wright respectfully. Jerry can we start with you?

Jerry Festa
Sure, how are you today?
 
John McHale
Great

Jerry Festa
I work for Curtiss-Wright Defense Solutions in part of the aerospace instrumentation group. This facility is in Newtown, Pennsylvania along with our sister division in Dublin Ireland. We've been here about 16 years and focus basically on the aerospace industry, however the last 12 years we've expanded in the space industry with our instrumentation product line. I'm the Product Line Manager for Space Systems and I've been tasked with growing that portion of the business over the last 12 years.

We provide instrumentation during developmental flights in space. Sometimes we are able to convince the customer that the reliability is there to also be used for the operational flights and that's been happening more and more so lately.

John McHale
Thanks. Mike?

Mike Deliman
Hello all. I'm Mike Dalaman. I've been with Wind River for over 25 years. Currently I mostly do instruction for the various VxWorks Wind River packages but I've done everything from tech support to custom architecture reporting including assisting the team that did the RAD6000 port back in 1995 for the Mars Pathfinder - that's that's what put us into space.

John McHale
Thanks Mike - now the RAD6000 that definitely was not a COTS product. Would you agree?

Mike Deliman
Actually that was a COTS product getting made by the RAL Federal Space Systems - a division of IBM and it was the very first COTS product that went into space selected for the new Millennium missions - the Mars Pathfinder in specific and yes it would actually was a commercial-off-the-shelf board.
 
John McHale
And yet going right into our first topic of discussion - why is COTS in space kind of a taboo word among this industry and those folks there even though they had products like the RAD 6000 were going into space and they were off-the-shelf. Maybe they just didn't like the word commercial. Mike, since you kind of went there can you start us off on that?

Mike Deliman
Sure. So historically and even currently today when you think of commercial or commercial off-the-shelf the first thing you think of is stuff being shipped from Amazon you know there's already a million of them on the shelf, they've been shipped all over the place, maybe they're the most basic kind of feature sets, mass-manufactured so how could they actually handle the rigors of space? So as you think of COTS as in the space race you think about this is stuff that's going to be used terrestrial projects, stuff that's going to be in your stereo on your TV, whatever. You don't think of satellites or deep space. In reality it just means that it's being made for more general purpose rather than the traditional space applications where they may have built one or two or three systems and then the next project builds everything from the ground up again.

John McHale
Jerry, what about you? It's similar for you?

Jerry Festa

Yeah. the major issue has been reliability, and how you can quantify that. The statistical acceptance of the piece parts has been the issue for quite a while within the space electronics industry. The parts themselves - traceability issues and dye issues, the fact that most of the parts are plastic and not hermetically sealed. This has always been the issue with the systems that we provided in the past. So we've upscreen tested the systems at the system level environments to give the customer more confidence that the equipment can survive these environments.

Is it valid to up screen at a system level? Do you still need the parts traceability and data as far as the reliability for those specific piece parts? These questions are still there to some extent. We do provide a lot traceability. When required we do upscreen the piece parts , so there exists a happy medium  between COTS and military requirements. Prior to that there was a substantial market built around the military standards as well as the NASA space standards for parts and materials. The exorbitant cost of that is why COTS is now being looked at from a reliability standpoint not seen before.

John McHale
Jerry, quick follow-up to that. I mean is there also a cultural barrier to trying to educate folks who have been kind of subconsciously not wanting to go to the COTS route - is that part of it as well? The education challenge with the culture?

Jerry Festa
I think so, it's been entrenched in the culture for a substantially long period of time. A lot of the, shall we say, space pioneers are getting up there in age and there's a middle age group that is not there. There are either gray beards or there are fairly young engineers 5-10 years out of college. The big age gap and their different experiences has affected their acceptance of COTS. So I think that the information exchange that's happened between those different age groups and the extent of the testing that can be done to screen out any parts issues is better than it has been and that's helping with the acceptance of COTS.

John McHale
Thanks Jerry, and Mike you hate to say it but you're a bit of a gray beard if you were working on some of those early programs but maybe you can speak to that a little bit. Were there examples of, you know we talked about trusting the COTS in the culture but were there any failures that happened and perhaps didn't get much press and may have made the industry a little more hesitant to try this model until the, you know, obviously the cost pressures changed.

Mike Deliman
As a graybeard - and yes I actually wear that moniker proudly - I've been part part of the space industry from representing Wind River and working with guys at a couple of different manufacturers of space boards if you will through from being one of those young guys all the way now to being a gray beard, although there were some projects that failed - most notably a couple that went to Mars at the end of the 1990s. They really weren't from the hardware itself failing it was more overview of actual engineering processes that sort of failed. So the COTS hardware was cleared in these problems and that partially helps give COTS a better name for space.

So yeah there were failures but it wasn't the fault of the parts themselves. So in looking more at how COTS came into play, before we started using COTS in space, just looking at Mars itself maybe 35% of launches actually succeeded, actually not only got off the pad but got to Mars and did what they were supposed to. Since 1996 with that first COTS project - Mars Pathfinder - something like eighty percent of the launches have not only succeeded but gone on with multiple extensions running into the tens and almost twenty years longer than they were designed to run.

John McHale
Jerry, how about you do - you have any experiences like that? I mean I remember attending conferences maybe 20 years ago where people were talking about upscreening and I see some of these folks in the satellite industry just start rolling their eyes but yet it was being done because they want to get the product up there. I mean, do you have any anecdotes?

Jerry Festa
Well, just to clarify with Mike, I am a graybeard as well. I started in 1980 with RCA Astro and launching satellites out of Vandenberg Air Force Base for the Air Force, so I've been through a lot. Been through some failures out there. The the traceability for the failure, doing the failure analysis of a launch failure is extensive and goes all the way down to the piece part level and there have been instances where parts have failed or at least blamed for the failure. To think that you don't hear much about a lot of the military failures and what actually caused them - they don't broadcast that too much but there have been a few.

John McHale
Yeah, I don't get press releases on those for some reason.

Jerry Festa
Yeah and I think that over the years it's kind of been cyclical. I mean we've gone from never using COTS to sometimes using it and then there'd be an issue and that it would kind of scale up the other way and then over a period of time when you have cost issues or schedule issues COTS has been utilized and someone, you know, will for the reliability at the system level or use redundancy instead and use different manufacturers for those same redundant parts so that you don't come off the same line so the reliability numbers work out for you as a system engineer.

So there are a lot of ways to get around it but I think that it's definitely cyclical and then if it does happen again and they they realize, hey there's an issue with a part, the industry handles it and I think that the production and the lot-traceability and the process controls - all those things have been increased, the quality of those processes has been increased to the point where it's not how you build the part but you have to be able to prove that you built it to those those processes and that traceability through the process for the manufacturing has gotten a lot better for COTS.

John McHale
And thanks Jerry - that kind of brings me to the next question. So what's changing - we're hearing so much more about COTS being used in space and I'm going to go with you and I feel free to mention you guys have, well Curtiss-Wright is kind of bringing a COTS solution to many space programs and you can talk about that and then Mike if you could follow up as - take it maybe from a software perspective, what's changing.

Jerry Festa
Yeah, I think that what's really changed is this move to LEO and the mission life are much lower and the radiation requirements are not as extreme as the long term missions up in GEO. So now the military is looking at LEO just like the commercial companies have today. Now where we're in the middle of it because we provide data acquisition systems for these launch vehicles that are sending everything up to LEO and they're usually first and second stage. The first stage will use the same hardware but we use a radiation tolerant version of it for the second stage which includes what's called a smart backplane.

The issue with the second stage is typically that the radiation requirements are higher and you have to protect against what's called single event upsets (SEU). Curtiss Wright has developed a monitoring backplane for all of our data acquisition stacks that monitors the current to any of the modules in that stack. If you do experience a SEU, the current typically spikes and it will hold up at that high level until you reset the power to that unit. And we can automatically sense that as well as automatically reset the power. It’s a way of providing Radiation Tolerance against SEUs which is the major issue with any electronics equipment in space.

The electronics can survive the total dose for a five-year mission these days with COTS equipment but the upsets are still there. They happen periodically. Some of our tests have indicated more than a hundred upsets over an accelerated life test equivalent of a five-year span in Leo. So you need to reset the electronics, otherwise that high-power will eventually cause issues with the card or the module. The resets allow the use of the particular equipment for the five-year span, which is typical for LEO missions.

Especially with the small satellite constellations - they're all being designed for five-year life. They're smaller, cheaper and easier and less expensive to get into orbit. Instead of spending a billion dollars for a GEO spacecraft you can put up a constellation for the same amount of money that covers the entire globe and can be replaced every five years as the technology evolves.

John McHale
You told me very recently, Jerry, that when you have that model, that's a peak part of the model that's changing, the small sets. Where before they design for almost a fifteen-year life, now it's like five to seven and I think what you said is if someone, i think the quote you said was 'if someone's building a satellite today, the last 15 years, the tech up there is going to be obsolete in five to seven with other satellites going up' or something along that line.

Jerry Festa
That's the thought. I mean, in any technology on the ground there's a tech refresh. In the aircraft industry there's a tech refresh cycle and it's five years because five years is about when the technology is at a point where it makes sense to swap it out. So we swap out boxes in aircraft every five years in some of our agreements with our other customers for aircraft and the same is being considered for spacecraft and the five-year life fits right into into our tech recycle and most technology the way it's evolving.


John McHale
Thank you Jerry. And Mike, what about you at software side - what are you seeing there? Is there any, what's changing on that end? Has it been being adopted a little bit longer maybe than some hardware?

Mike Deliman
Yeah, this is going to be an interesting set of answers. So just expanding and explaining some things, Leo is Low Earth Orbit, geo is Geosynchronous Earth orbit. These both take advantage of the Earth's magnetic field for some shielding whereas the stuff we do with NASA tends to go outside the Earth's magnetic field into deep space and has to endure much higher radiation levels. But having said that, so COTS and this increased lower orbit basically means lower costs and higher availability so instead of having to design a system every time you go to launch you can go and grab something off the shelf and implement on it and one of the benefits you have is this technology refresh and this inspection into failures making more robust hardware.

The next thing you get is since you're using an architecture that's already been used for space you can start developing libraries of software so the libraries of software are bits of software that you reuse and reuse and the more you reuse software the more mature it gets, the more highly debugged it gets the more tightly tailored it gets for the actual deployment you're putting it into. So this makes it so you can actually push software from one set of lines to the next set of lines standardized on a specific operating system or family of operating systems and development processes and that in turn gives you this extra maturity, this extra robustness on the software end.

In addition to that since the nineteen, early 1990s, we've had more ubiquitous use of things like DO-178C or ARINC 653 which are designed for flight systems and allow you to increase the robustness of a system by protecting applications from each other. So there's a lot of stuff that happens all in parallel over the last 20 years that just make it easier to create a very robust system and get it into space and and taking a step again towards towards the small SATs or cube SATs,  the paradigm is exactly that you now can develop a satellite for tens of thousands of dollars instead of hundreds of thousands of dollars and it's going to take more advantage of recent developments where, when we talk about the deep space things, the development cycle is more like 10 to 15 years so you're not going to get a huge jump in capability for 10 to 15 years so you can take advantage of the newer hardware and newer developments.

But back to it - what we're seeing a big jump forward in is there's a lot more things going up. a lot more projects going forward and a lot more players getting involved and from my end of things the best thing is that maturity of the software libraries that get reused and move forward along to help enable the space guys to do their stuff.


John McHale
Thanks guys. So we talked about what they have to do and a lot more players coming into the market. We had a guest on a previous episode, Tony Jordan, talk about how a lot of companies are kind of coming up with a new set of product lines to kind of meet a demand for what Jerry described as these products that are RAD hard enough but not have to be super RAD hard and can meet the cost, basically the cost requirements. I think it's going from multi-million dollar payloads down to three hundred thousand dollar payloads or three thousand dollar satellites so it's a big challenge. There's a business challenge on the providers - what is there any challenges, Jerry, remaining that you think, sorry, challenges and drawbacks to using COTS that your hearing from your customers or challenges that you guys have to overcome in this area or is it more about just getting and proving your product in space?

Jerry Festa
Well I think when you talk about mission critical systems the reliability needs to be there for life-threatening, let's say, failures so that's when you start moving into more and more higher-critical systems COTS, the reservations with COTS are still there, so the systems designers will typically use COTS but they'll end up having to to get the reliability numbers up, they'll have to design in redundancy, let's say, or upscreening the halt for all the parts. That all adds additional cost so you can actually get to the point where, well, we're back where we were with the mil-standard space qualified hardware but I think as they get more and more comfortable with using upscreen COTS for these critical systems where you have redundancy built-in, I can see down the line that evolving to the point where maybe the redundancy is no longer required because you have flight heritage now so that you've flown enough of these these redundant systems that, without failure or switching over to the other side, so that you you may want to rethink your reliability calculations and maybe you could fly with just one

But I think that the reservations and the drawbacks I think are still there as far as screen parts - we've got to find a way to decrease the cost for upscreen parts somehow. That seems like that's an area where you could really save time and money going forward.

John McHale
And I would think with some of those you mentioned - the military in respect to the old mil  spec parts so the military's embracing COTS in space maybe not as much as the satellite providers, I mean maybe not for the classified missions but you're still seeing some kind of change there at all or more embracing of it?

Yeah, somewhat, somewhat. I mean when you think of commercial satellites and military satellites, they have the same reliability requirements that they're putting up an expensive piece of hardware and they want to make sure that it survives, especially with the the Geo spacecraft that have to survive for 15 years and most of the the old comm SATs are all up there. Like when I was with RCA, it's now like more the A2100 is the standard bus that's been used for the past 20 years now by them and it survives 15 years. It was designed to survive for seven and when they were surviving for 15 it almost put the company out of out of the space business.

But I think that I think we're we're in a pretty good spot now where the commercial folks are kind of leading this charge into COTS that the government is now saying there's something there - we want to save that type of money as well - let's get our our systems guys analyzing what the what the issues are as far as reliability and risk mitigation, what we can do on our part and well, I think we'll end up at the same place down the line.

John McHale
Thanks. And Mike, I know you've worked with some NASA programs before - they're kind of embracing the use of COTS products in space and but and along those lines can you speak to that and maybe talk about any remaining challenges for the software guys

Mike Deliman
Well I'll look from the hardware point of view first. Again you still have the availability of hardware and the cost of it and especially when you get into the deep space stuff you have the additional testing and and screening of parts so on and so forth. And then once you have the hardware qualified there's still not a whole lot of it sitting on the shelf waiting to be used. we can address some of that with simulators - things like the Simex simulators that actually mimic the entire board right down to the CPU and instruction set. Those can give a jump forward in development and then we have SWAP - the space weight and power usage - the computers still have a physical amount of weight and power that they have to have in order to provide the capabilities that we need so those are still kind of at a premium.

There's never going to be software fast enough on the hardware to make the scientists and computation experts happy. We're always going to need that that responsiveness, the real-time determinism of software, to get the most out of that hardware. The development cycles are being more and more compressed. The need for having robust software is still very high so I think from a software point of view, the big problems are actually having hardware to run on. The development cycle is becoming more and more compressed just like we see in the commercial space and the the ways to address that are better simulators and more robust better-tested libraries of software.

John McHale
Thanks Mike. I appreciate that. Next up we're kind of on our final topic. We talked a lot about commercial companies, more companies getting into this world because of the small satellites. They're driving innovation there. They're inviting more people to play in this market, more COTS opportunities but the space markets always been kind of difficult to crack. What would you say are the biggest barriers of entries maybe for new COTS suppliers in the space electronics market. Jerry, I'll give you this one.

Yeah, I live with that every day, or I did initially. Flight heritage for space hardware providers is a box you have to check. So how do you get spaceflight heritage? Well NASA has a program that allows you to fly on test beds every now and then, which a small fledgling space technology company should should evaluate that and possibly get there that way. You get your heritage that way. Now if you have to overcome a number of things, I mean there's safety margin analysis that has to be done on your design reliability predictions. The parts traceability and the processes used in your production have to be there so it's only if you're a commercial house and you want to get into space, if you're not careful you'll end up having two production lines. Then you have to find a way to utilize your existing product line, manufacturing line, production line, I'm sorry, and institute some changes that can get you over that hump with space requirements and in the long run it helps you out in your commercial business because it makes your processes higher quality, more reliability, your equipment becomes better, you don't have the failures you typically will have over time.

So that transition from a production line for commercial products to space products is something that to maintain the low cost thatyour COTS equipment has you have to tweak your commercial production line essentially. So it's hard to do - it takes some expertise - you've got to hire additional people that have that expertise so there's some barriers to entry from experience, as well as from the cost, and small fledgling companies typically don't have that luxury and that's a barrier.

John McHale
Mike, what do you see from your perspective?

Mike Deliman
Almost the same exact kind of things. We had sort of a leg up with the Pathfinder project in that we had already been working with JPL for many, many years. I want to say JPL was something like Wind River license number four. So we had already been working with them - they used our software in their test beds so they knew something about the software - that kind of made it easier for us to get in. So from the software point of view you have to have that - that legacy - do you have any successful deployments? Where were those deployments? How did they work out? What sort of advantages did you give to the overall development team both for the hardware and and for the software end?

So with today's things we're getting more emphasis on prototyping - really fast prototyping - and modeling languages generating code. How can you be sure that that code is going to be running on a platform that gives you that extra push forward in development and cost savings. It's sort of reducing that time-to-market thing, kind of like in the commercial world, you want to be able to make sure you're on schedule and your stuff is going to go up on time to make your your mission succeed. So it's kind of kind of similar to what you see in the the hardware market in that unless you've already been there for a while and had this legacy of success and have enough resources behind you to provide responsiveness to your customer, you're going to have an enormous climb uphill just to get into the market

John McHale
Thanks Mike and thank both of you for an outstanding discussion. Learned a lot today and I think we can safely say COTS in space is here to stay, especially based on the success that Curtiss-Wright and Wind River are having in the industry and when we continue to see. And, you know, so also like to tell the audience a little bit about Curtiss-Wright Defense Solutions - if you want to learn more about their space products you can visit curtisswrightds.com/products/space/ and please be sure to check out our next episode on open architectures in space. Just visit windriver.com/new-space-race. That's windriver.com/new-space-race. I'll see you there in the next broadcast.

Jerry Festa

Author’s Biography

Jerry Festa

Director of New Business-Space

Jerry Festa is a Program Manager and the Space Product Line Manager for the Curtiss-Wright Teletronics and Acra Business Lines. Jerry holds a BS in Electrical Engineering from Drexel Univ. with MBA studies at Drexel and Arizona State Univ. He has over 25 years of Space Industry experience in Program Management, Systems Engineering and Business Development.

Share This Article

  • Share on Linkedin
  • Share on Twitter
  • Share on Facebook
  • Share on Google+
Want to add a comment? Please login
Loading...
Connect With Curtiss-Wright Connect With Curtiss-Wright Connect With Curtiss-Wright
Sales

CONTACT SALES

Contact our sales team today to learn more about our products and services.

YOUR LOCATION

PRODUCT INFORMATION

Support

GET SUPPORT

Our support team can help answer your questions - contact us today.

REQUEST TYPE

SELECT BY

SELECT Topic