“Yes We Can”: Building Hybrid Rugged Systems Using a Mix of Applicable Embedded Standards
Why making “products” instead of making “projects” is catching on for SFF and other manufacturers in the military and aerospace markets.
I am sure many of you have heard the expressions,” You can’t have the best of both worlds” or “You can’t have your cake and eat it too.” But with all the modernity and technology, I ask myself why can’t we have the best of both worlds and have our cake and eat it too? Designing a Hybrid standards-based computer offers an optimal solution—one stronger than just relying on a single dedicated standard or a proprietary solution.
Historically, engineers would design and build a computer exactly to the specifications of a project. If the parameters were rigid and never expected to change, the engineers and Project Manager could choose the easiest path, taking costs into consideration. Both a single standard based or proprietary solutions had their advantages. With standards, there are more options available in the market, with easier integration if fully compliant to the standard, and shorter development time.
However, if the customer using standards needs changes to the available product, NRE costs start to become substantial. Proprietary solutions, on the other hand, deliver a level of control that enables engineers to get exactly what they want without compromise. On the down side, the proprietary approach eats up man-power and means longer times for deployment and compatibility challenges, which can lead to higher costs in the long run. A key disadvantage of both strategies is seriously limited flexibility. Employing a single standard only allows what the standard can support and lowers the number of suppliers. Choosing a proprietary solution offers the least flexibility, as the outcome is only relevant for that project. Making any minor changes would require a redesign from scratch.
Today, many of the major military and avionic manufacturing companies are changing their strategy. They are now more interested in making “products” instead of making “projects.” Products must be flexible to meet specific end customer requirements. There is no such thing as “one product fits all.” Flexibility enables end customers to choose what they want, when they want it. This means the component capabilities must be swappable at any time and not permanently soldered in. Products must be flexible to successfully work with processing power differences, with the degree of I/O support, and with different network types including MIL-STD-1553, ARINC-429, AFDX, TTP, and Ethernet. Flexibility is key as well to accommodating video requirements such as video I/O quantity and formats, encoding, video capture and so on. Some products, seeking flexibility, will require FPGAs to allow for application-specific, high-speed data requirements. Achieving flexibility isn’t the only challenge. Now manufacturers face additional demands for these products to be as small and low in weight as possible but powerful enough to run the applications with the least amount of power. When competing at the product level, price becomes a key differentiator, thus driving prices downwards. The term that everyone knows as SWaP-C summarizes all these new challenges.
It will be very difficult to achieve these new demands by just relying on single standard solutions and/or proprietary ones. Single standard solutions, with greater flexibility than proprietary approaches offer, take manufacturers in the right direction. However, standard solutions can fall short. For example, 3U VPX is a well-accepted small form factor standard and allows for powerful computing. However, if a product only needs limited functionality where two to three 3U VPX boards are needed, the computer loses its SWaP-C advantages. Proprietary could be a better response to that but capitulates on flexibility. Combining proprietary and standard can capitalize on the strengths of both while offsetting the disadvantages.
Hybrid Solution-Based Explained
The basic concept of the hybrid solution is to have flexibility throughout the building blocks. The first of the building blocks is at the component level. CPUs, GPUs and FPGAs are adding more capabilities by merging lots of functionality onto a single chip. System on a Chip (SoC) components are now available on many standard small form factors such as XMC, COMe types and MiniPCIe. The second key building block which supports the variety of standard based connectors is the midplane.
The midplane is crucial, as it must be designed to support as many different types of standard connections as possible while minimizing its footprint. For example, the midplane should use both sides of the board. It is important to pay attention to thermal management; thus placement of the connectors’ sites and implementing the cooling mechanism is imperative. The next building block is the I/O Transition Board, where all the I/O is routed to one side, and other side is used for the appropriate high-density, high-frequency Mil-style connectors.
The outer building block is the rugged enclosure. It is broken into three sections. The midsection is the central chassis body, and is designed to be reused without modification. This feature will preserve standardization where there is no need for NRE, thus keeping costs down. In its standard configuration, it is equipped with captive hardware for bolting to a mounting shelf or bulkhead. The System Front Panel section is designed to bring out all available I/O in its standard configuration. It should be easy to modify for a specific application. The Chassis Rear Panel section should be designed to bring out I/O from standard boards that have “Front Panels” (such as WiFi, GPS and SDR). An intelligent designed skeleton system using a hybrid of standard connectors will promise the ultimate level of flexibility while adhering to SWaP-C requirements.
Since the connectors are based on standards and the midplane is passive to receive them, no special drivers would have to be created.
One of the oft-observed traps that COTS manufacturers fall into is a tendency to make systems that are solely based on their boards or optimized for a specific project. The computer becomes limited in its design, even though it’s based on standards. In other words, the manufacturer is “building-out” from their products only. To be successful in making a flexible solution that can be reused in many projects and products, the system should be designed by “building-in.” Building-in puts the engineering effort in the chassis and midplane, thus allowing more flexibility when choosing COTS payload modules. The system should not be limited to any single COTS vendor.
More Flexibility, More Opportunities
Hybrid rugged systems can now open up to more opportunities by adding more suppliers to the market that will lead to lower prices and overall costs. There are many companies, or departments in large organizations, that developed a special board in a standard form factor like XMC or PMC or COMe, but they don’t have resources to build a system. The advantages of a hybrid system that is based on universally available standards is that it can be converted into a Hosting Platform. The hybrid system would come in a skeleton form or include the requisite COMe based CPU, SATA storage and relevant I/O; the company or department can simply place its standard based board in the available slot and immediately have a fully rugged working system.
With all this flexibility, a hybrid system can be applicable for military, avionic and heavy industrial products and projects. This approach promises to be far more sensible and cost effective over relying on a single vendor for all the modules, a path that can lead to amassing heavy development and NRE costs.
In conclusion, building Hybrid Rugged Systems using a mix of applicable embedded standards offers a lot more flexibility to engineers and project managers for their products and projects. The Hybrid Rugged System approach offers more supplier options, which in turn lowers costs and makes it easier for engineers to find and integrate exactly what they are looking for.
So, to answer the question: “Can we can get the best of both worlds?” AND “Can have our cake and eat it too?”, I simply reply “Yes We Can!”
Pawan Seth is a Director at Alligator Designs. Alligator Designs is an AS9100C Certified Company providing world class rugged product and system designs, developed and deployed on land, at sea, and in air for close to 15 years. Visit www.alligatordesigns.com or email email@example.com, and receive further information about the company’s latest Hybrid SFF computer, Falcon II.