COMPUTER-AIDED DESIGN TECHNIQUES FOR AUTOMATED TEST PROGRAM DEVELOPMENT

THE DEVELOPMENT OF COMPUTER-AIDED DESIGN TECHNIQUES FOR AUTOMATED TEST PROGRAM DEVELOPMENT IS PROPOSED. THE APPROACH WILL BE TO VIEW THE TEST PROGRAM DEVELOPMENT PROCESS AS CONSISTING OF EQUIPMENT MODELING, FUNCTIONAL TEST SEQUENCING, TEST PATTERN GENERATION AND SEQUENCING, TEST VERIFICATION, AND AUTOMATIC CODE GENERATION. OF THESE, TEST SEQUENCING AND TEST GENERATION CONSTITUTE THE TWO MOST TIME-CONSUMING ACTIVITIES. THE IMMEDIATE RESEARCH WILL ADDRESS THE DEVELOPMENT OF COMPUTATIONALLY EFFICIENT TEST SEQUENCING AND TEST GENERATION ALGORITHMS USING ADVANCED TECHNIQUES FROM INFORMATION THEORY, ARTIFICIAL INTELLIGENCE, AND MATHEMATICAL PROGRAMMING TO SOLVE VARIOUS FACETS OF THE PROBLEM. THE LONG-TERM GOAL IS TO INTEGRATE TEST SEQUENCING AND GENERATION ALGORITHMS, OTHER ASPECTS OF THE TEST PROGRAM DEVELOPMENT PROCESS, AND USER-FRIENDLY INTERFACES INTO A COMPUTER SOFTWARE PACKAGE FOR USE BY TEST PROGRAM DEVELOPERS AND AUTOMATIC TEST EQUIPMENT (ATE) END-USERS. THE PHASE I RESEARCH EFFORT WILL DEVELOP INFORMATION-THEORETIC/AI ALGORITHMS FOR TEST SEQUENCING (TO MINIMIZE THE COST OF FAULT ISOLATION FOR DIGITAL, ANALOG, AND HYBRID MODULES) AND MATHEMATICAL PROGRAMMING TECHNIQUES FOR DIGITAL TEST GENERATION TO DEMONSTRATE THE FEASIBILITY OF THE APPROACH. THE POTENTIAL BENEFITS INCLUDE COST-EFFECTIVE TEST PROGRAM DEVELOPMENT, MINIMIZATION OF EXPECTED TEST TIME AND FALSE ALARMS, BETTER ATE-HUMAN INTERFACES, IMPROVED MAINTENANCE THROUGHPUT VIA THE ABILITY TO ADAPTIVELY SELECT TESTS BASED UPON ARCHIVAL FAILURE DATA, AND REDUCTION IN THE NUMBER OF ATE SYSTEMS NEEDED.