A CASE STUDY
Edward DePalma
Rome Laboratory
ABSTRACT
Transitioning military technology to commercial applications is an important step toward cutting the costs that have grown from maintaining military specific processes and procedures. Rome Laboratory has accomplished such a transition in an area of system development that has long been identified as high cost and labor intensive. Military standards of the past mandated that detailed tasks and techniques must be applied during the development of military systems to ensure their reliability. Often these mandates came at a high cost and had a severe impact on development schedules. This resulted in a choice between trimming these tasks and losing whatever benefit they may have had or perform the tasks and go into costly budget and schedule overruns. One of the most costly and poorly performed of these tasks was sneak circuit analysis (SCA). SCA searches for system design flaws that produce unintended control paths that will inhibit a desired function or enable an unintended function. Since these flaws are designed into a system and cannot be identified by stressing the system, they evade many reliability tests.
The analysis itself is a detailed schematic level search for design flaws that requires a knowledge of performance specifications, system design, and any interfaces that exist with outside systems. To be effective, this analysis must be performed on final designs since any subsequent changes could introduce new sneak conditions. The problem with this approach is the lack of time between setting the final design and the start of system production in which to complete such a detailed analysis. The result of this conflict is the analysis being performed after the production has begun with any flaws that exist being designed into the system and possibly corrected later at greatly increased cost.
A solution to this problem is the automation of the analysis, performing it throughout the design phase of system development, and taking the corrective action before production begins. Rome laboratory has developed such a tool (RL/CAPFAST / SCAT ™, Sneak Circuit Analysis Technique), and has transitioned this software out to the commercial world where it has been applied to both military and commercial designs. The RL/CAPFAST / SCAT ™ software is currently being marketed by the Phase III Logic Corporation and has been integrated into their schematic capture program CapFast where it can be called on to search for, and identify, sneak circuits. This paper focuses on the software tool development, the successful transition to a commercial vendor, and the current military and commercial applications.
Note: the paper was presented at NAECON in July 1997 in Dayton, Ohio.
Please contact Dr. Chong Lee for more information