TRL: Putting It in Perspective
In the DoD’s quest to further leverage commercial off-the-shelf (COTS) technology and the maturity it offers toward faster and more affordable deployment, they often demand high TRL (technology readiness level) early in the program stage of development. TRL defines technology readiness levels from 1 to 9—1 being basic technology research, and 9 being defined as application of the technology in a deployed operational environment. New program solicitations and developments typically mandate high TRL levels at the inception unless new basic technology ground needs to be explored.
It is typical in the defense electronics space that the final embedded computing solution qualification and deployment is a couple years or more down the road from the start of engineering and manufacturing development—yet high TRL is demanded for baseline computing product at the outset. The DoD and system integrators should be less hung up on TRL when considering industry standard embedded computing technology for systems that are a few short years away from deployment, and more focused on the underlying capabilities of the open standards embedded computing ecosystem and its foundational technology.
Managing TRL is about understanding when, and at what risk, to integrate new technology into programs—the primary objective being managing risk and controlling cost. And if that’s the goal—the fact is that the open standards embedded computing eco-system is built on robust proven standards with fairly predictable outcomes, especially a few years out, and can be relied on to be both minimal risk and low cost. The foundation and maturity of the embedded computing technologies piece parts support an argument that future, yet-to-be-designed products can be considered to have a high TRL—perhaps even TRL 6, which is defined as a system model or prototype demonstration in a relevant environment—before the product is even conceived on paper. That is, so long as the new designs are not breaking new "technology" ground.
Like buying a car
I suppose one could use the analogy of buying a new car and perhaps making a decision on a model before it is even off the draftsman’s drawing table. All the expected piece parts that need to make this car suitable for its intended purpose are fairly mature (high TRLs) save for incremental improvements; tires are well characterized, internal combustion engines evolve successfully, chassis don’t demand too much innovation and cabin safety, control and comfort features are near commodity. There’s pretty much no risk that it won’t work as you expect, or that it won’t be available when promised.
It’s the same for the open standard embedded computing industry; silicon technology development and manufacturing processes are mature and refined, and the same common backplane connector technology and parts are used across multiple applications and products. Form factors, particularly Eurocard, have been around for the better part of 30 years in defense electronics with tens of thousands of products deployed and long deployment life cycles providing significant empirical evidence of maturity, or readiness.
New embedded computing product designs are—in the simplest terms again—just evolutions from previous high TRL designs, but deliver new capabilities, features and performance. We can argue that a design that is necessary to meet future requirements in two to three years has TRL 6 qualities and therefore can be a baseline for future defense computing applications with little risk.
Mandating high TRLs is less relevant with defense computing where industry standard products are used. Extrapolation to the end solution poses minimal risk. That’s the beauty of COTS and open industry standards. It’ll be ready when the time comes.