Smart BIT-2: adding intelligence to built-in-test

It is noted that the current high levels of false alarms, difficulties in properly identifying intermittent failure behavior, and the integration of environmental stress data into the BIT (built-in-test) decision process are important characteristics that call for improvement. Smart BIT-2 addresses these problems by the application of artificial intelligence and related techniques to the BIT decision process. The results in three new smart BIT approaches: adaptive BIT, temporal monitoring BIT, and opportunistic diagnostic BIT. The first two of these were demonstrated/validated using testbed scenarios. The resulting system showed the capability to discriminate against the typical false alarms that plague conventional BIT systems in the field. A software overview is given, followed by a discussion of the empirical results