Designers face the downside of COTS - increased risk

HILTON HEAD, S.C. - Military and aerospace electronics designers are starting to digest some of the brute facts involved with the switch to commercial off-the-shelf (COTS) components. Chief among these sobering realizations is something they all hate - risk.

By John Keller

HILTON HEAD, S.C. - Military and aerospace electronics designers are starting to digest some of the brute facts involved with the switch to commercial off-the-shelf (COTS) components. Chief among these sobering realizations is something they all hate - risk.

Military systems designers traditionally take the conservative approach. When a piece of military equipment is going to the field, not just to the research laboratory, the designer wants little risk and dependable incremental improvements in system performance.

Commercial systems designers - the inspirations for COTS - lean toward the big-risk/big-potential-payoff approach as a means to steal a march on the competition. This, however, is anathema to the military designer`s core beliefs.

In the best of circumstances, the military designer wants the best-quality parts available - no matter what the cost - to minimize physical risks to the end users. To do this, they seek known quantities - tried-and-true parts with documentable pedigrees that they know will work under a wide variety of conditions.

Yet the rise of COTS signifies the decline of military specifications, standards, and test procedures, which are the means by which system designers understand component reliability.

"Off-the-shelf standard microcircuits do not exist. There are no standard plastic parts," points out Bahig Tawfellos, chair of the Electronic Industries Association Task Group 97-03, and a designer with the AlliedSignal Aerospace Government Electronic Systems division in Teterboro, N.J.

The risk of COTS stems from the unknowns that this new design approach represents - unknown component performance under extreme environmental conditions, unknown long-term vendor support, unknown performance after long-term storage, and unknown reliability test data among separate vendors.

"Commercial RF parts are a big problem because of their variability," says Brian Willoughby, of the Computer Sciences Corp. Best Manufacturing Practices center of Excellence in College Park, Md.

Adds James Jorgenson of the Sandia National Laboratory Electronics Quality/ Reliability Center in Albuquerque, N.M., "The long-term dormant storage environment is unaddressed by commercial industry." This, he points out, is of particular concern to nuclear weapons designers whose systems must remain in storage for decades without suffering any compromise in reliability.

The COTS movement "changes everything. It can change things for good or ill," admits U.S. Navy Capt. William Shotts, commander of the Naval Surface Warfare Center (NSWC) in Crane, Ind. Shotts and Tawfellos made their comments in November at the NSWC`s Technology Management Symposium & Expo in Hilton Head, N.C.

Granted, using COTS parts can enable designers to develop systems sooner and at less cost than they can with mil-spec parts, and can use rapid COTS technology insertion to keep military systems up to date, Shotts says.

Yet at the same time, rapid technology insertion creates higher risk over time than do mil-spec designs because systems are constantly new and thus have immature test histories, Shotts says.

More in Test