SOSA initiative gaining momentum in defense electronics community

Story

January 20, 2020

The U.S Air Force hosted the FACE and SOSA Expo and Technical Interchange Meeting (TIM) event in Dayton, Ohio during September 2019
“to see the progress being made for both functionality and interoperability of modular open system environments in both the Future Airborne Capability Environment (FACE) and the Sensor Open System Architecture (SOSA) consortia,” which are managed by The Open Group. The roundtable below consists of members of the SOSA consortium who exhibited at the TIM. The panelists discuss the effects SOSA is having on the military end user; how integrators, primes, and the Department of Defense (DoD) are embracing open architectures; and how to ensure that such
added structure does not stifle future innovation.

The panelists are Paul Mesibov, vice president of engineering, Pentek; Greg Powers, global market leader, Aerospace & Defense Performance
Solutions Division, W.L. Gore; Mark Grovak, director, Avionics Business Development, Curtiss-Wright Defense Solutions; Michael Munroe, principal backplane architect, Elma Electronic; and Mark Littlefield, vertical product manager for Defense, Kontron.

MILITARY EMBEDDED SYSTEMS: The FACE/SOSA Expo and TIM was held in September in Dayton, Ohio, near Wright-Patterson Air Force
Base. What trends supporting the SOSA initiative did you see emerging at the event?

MESIBOV: As both a SOSA working group participant and an exhibitor at the TIM, we have seen the interest in SOSA really take off within the community of our customers. Many questions were asked about the readiness of the emerging [technical] standard and how it will impact the embedded systems market. If any trend can be identified at this point it is that the message to learn about and ultimately include SOSA as part of our customers’ solutions is coming from high up in their organizations.

POWERS: There are multiple  OSA trends materializing now, including a growing list of participants that represent the ecosystem, the emergence of products and systems targeted for SOSA conformance, the spooling up of the SOSA Outreach Committee and coherent messaging to industry, and the inclusion of anticipated SOSA conformance into contract language. Each of these trends points to the building momentum of the
standard, its maturity and the ecosystem rising as intended.

GROVAK: It’s evident that SOSA is gaining acceptance by suppliers who see the success of FACE and want to get in early. Companies are looking for early definition of SOSA requirements so that they can incorporate them into their product roadmaps.

MUNROE: There were multiple chassis using the CWECC Timing Card and a variety of SOSA- and CMOSS-conformant development chassis. 25GBase-KR capability was implemented in two backplanes at the event, but it was not clear which slot and module profiles are already
supporting this protocol. Expansion plane switching appears to be an important new architectural feature.

LITTLEFIELD: First off, I was really surprised by the attendance. Just the sheer number of people interested in FACE and SOSA was a pleasant surprise, and the depth of the questions we got told us that people are serious about embracing these standards quickly. Regarding trends, it’s clear that people want hardware to work with now and are looking for performance. There were lots of questions about 40-Gigabit Ethernet and 100-Gigabit Ethernet roadmaps. There were also a lot of folks looking for [single-board computers], switches, receivers, FPGA computing, and storage solutions. [All this] tells me that people want to build real systems right now.

MILITARY EMBEDDED SYSTEMS: Why is SOSA so important to the military end user, the prime contractors, and the embedded hardware and
software providers?

MESIBOV: The primary goals of the military end user are to be able to procure tactical systems that are cost-effective to buy, can be delivered quickly, and can be efficiently upgraded or repurposed. The technical challenge in meeting these goals for the prime contractors and their supplier community is to make the methods and techniques used to build these systems more systematic without reducing technical capability.

POWERS: This is really the essence of the SOSA initiative and underscores the merits of the open architecture concept. The end users will experience a broadened supply base, increased alternatives of compliant systems and components, and improved platform and system speed to market as producers have firm configuration understanding for their products. Also to come will be accelerated capability evolution as participants bring new ideas and technology to the ecosystem.
Reduced development costs and gestation times should drive the DoD’s ability to fund more and more varied programs for an enhanced “warfighter toolbox.” Primes will see reduced platform time to market, resulting in faster revenue generation by having a menu of existing qualified solution sets that can be rapidly integrated. Hardware and software providers will have firm technical targets for their products and a waiting marketplace for new and MRO [maintenance, repair, and overhaul] products, providing confidence for portfolio investment and evolution.

GROVAK: SOSA provides a coordinated approach that takes advantage of a number of industry standards for quicker delivery of the latest technologies to the warfighter. By consolidating these standards, SOSA should make it easier for system integrators to have numerous suppliers compete, in order to provide the warfighters with enhanced capabilities in a shorter amount of time.

MUNROE: The SOSA extensions to VPX add a new level of interoperability together with critical new features such as radial clocks, RF, and optical backplane feed-thru capability as well as 25GBase-KR/100GBase-KR4 signaling capability and a push to support chassis and system management.

LITTLEFIELD: It’s really good that you differentiate between these four groups because SOSA is important to each for different reasons.  For the military end user there are simply two things of key importance – getting new systems fielded more quickly and enabling faster, more affordable upgrades of fielded systems. Both of these are aimed at giving the warfighter the best, most advanced tools to accomplish their mission.

For the prime contractors/integrators, SOSA holds the promise of an ever-growing stable of components from numerous vendors which have a reasonable expectation to work together out of the box  Not having to design vendor-specific slots means that one can more easily do a technology refresh without the burden of vendor lock. This will be even more true once the SOSA conformance process is in place and vendors begin to certify their products.

Now, at first glance SOSA might seem a bad idea for the vendor community as it appears to drive commoditization. While that is true to an extent, it will more importantly drive hardware providers to innovate in other ways in order to differentiate from their competitors. For software providers the value is simple – it creates a market where today there effectively is none.  It will pick up from FACE, VICTORY, MORA/VITA 49 VITA Radio Transport, and other open systems projects like COARPS and MAPS to bring a software component market to the high-performance sensor environment.  

MILITARY EMBEDDED SYSTEMS: What are the biggest hurdles for SOSA volunteers/participants to overcome in the next year?

MESIBOV: For the SOSA participants who are developing the technical standard, the biggest challenge over the next year will be to define what conformance to the standard means and how it will be verified at the product level. After that, as systems are built to the standard, it will be exciting to see it evolve as problems are found and fixed and revisions to the standard are published.

POWERS: The SOSA standard is well underway, but is still a work in progress, so there is some risk of specification drift when looking at first-generation products and systems. However, as with any new initiative, there will be a controlled ramp to market with continual learning and refinement. Concurrent with the drafting and adoption of the SOSA [technical] standard, steps in the path to market include concepting, prototypes, demonstrations, low-rate production, and a ramp to higher volumes as driven by proliferating DoD contract inclusion. A fundamental initiative accelerator that can’t be overstated is that SOSA is incorporating many existing standards, such as OpenVPX, Ethernet, and rugged connectivity via MIL connector specs, SAE AS6070, and ARINC 802. The integration of these proven open architecture building blocks benefits everyone in the ecosystem and further justifies participation in the tiers of standards.

GROVAK: A concern I have is that by defining hardware modules down to the individual pins, with no allowance for variation and also defining how the modules connect to each other, SOSA – while gaining some advantages from standardization and avoiding vendor lock – may also inadvertently restrict innovation by being so prescriptive.

An unintended consequence of SOSA’s drive to standardization may very well be the impediment and restriction of innovation. That’s because any
new leap-forward approach to integrating a system would not be allowed in a procurement process where SOSA conformance is required. That would put a company with an innovative approach between a rock and a hard place, because going to the SOSA Consortium in advance of a procurement to try and get that approach incorporated into the SOSA [technical] standard, would necessarily expose the innovation to competitors. On the other hand, not doing so would preclude that company from proposing their non-SOSA-conformant solution, even if it would be a more innovative and better solution for the warfighter.

MUNROE: A hurdle will be settling approaches and requirements for in-band (control plane) system management and the validation methods for all requirements and recommendations. Extending VITA 68 for 25GBase-KR signaling will likely be contentious.

LITTLEFIELD: So far, the SOSA community has been pretty successful at defining a constrained list of slot and module profiles along with implementation rules around them, and we’re seeing the first products designed against them hitting the streets. The big challenges for the next year will be twofold. First, we are attacking some of the more difficult topics like system management and standardized interaction and security policies for things like sanitization and non-volatile read-only memory (NVRAM) functionality. These things can be surprisingly complex and varied in their implementation.

Second, we are defining the exact conformance language for the various rules captured in the standard. We need to craft a conformance mechanism that is robust enough to ensure that the goals of SOSA are being met, while keeping it from being overly expensive or an undue burden to the suppliers and integrators undergoing the certification process.