APPLICATIONS

Superior Diagnostics

The Diagnostic Profiler run time uses a minimum set covering algorithm to identify the root cause of failure symptoms. This algorithm is very efficient because it uses both pass and fail symptom results. It provides meaningful information for all real world situations including out of order data, missing data, and next best test recommendations that take into account test order restrictions imposed by hardware or tester functionalities. The next best test recommendation can also be configured to give more consideration to tests which are less time consuming over test which are time consuming when both tests achieve the same diagnostic results. Failure rate information can also be used to configure the next best test algorithm to give more consideration to parts which fail frequently.

Cost Effective Model Development

Unlike other model-based approaches that require a full functional model of the equipment, the Diagnostic Profiler requires only the basic diagnostic information. It focuses engineering efforts on fault to symptom relationships. This approach greatly reduces the scope of model development. Unlike hard coded decision trees, the Diagnostic Profiler supports quick model adjustments that reduce the amount of work required to modify the diagnostics down to only the test execution code changes. This approach makes adjustments to diagnostics quick and easy. The Diagnostic Profiler includes several tools which can be used to validate and verify the diagnostic capability and support offline debugging. This reduces the resources needed to complete development.

Captures and Preserves Expert Knowledge

The Diagnostic Profiler provides the means to capture diagnostically relevant information about a system in a sustainable format. The model includes subject matter expert knowledge about the relationships of symptoms to faults. Once this information is captured, it can be deployed so that every technician has access to the smartest expert. With today's systems increasing in complexity, this capability is essential. For older systems where there are few experts left, it is an absolute must.