The iRel40 project consortium has prepared a third Newsletter (M24). Today, the reliability concepts are at the doorstep of major changes, moving from stress-based and knowledge-based to application-based approaches. This is strongly supported by the current development of machine learning, digital twin-supported diagnostics or prognostics, and health monitoring. The partners in iRel40 are paving the road towards these reliability approaches and drive new concepts based on the Physics of Degradation.
In this third iRel40 newsletter we present 15 selected project outcomes related to data science (e.g. artificial intelligence models applied to production data, machine learning approaches, anomaly detection), digitalization (e.g. digital twins), as well as new test methods and processes for driving reliability. We show results from semiconductor and IC design topics like conditional burn-in (#1) or reliability assessment in IC design projects (#4) to application-related topics like big data streaming applied to electro motors (#14) and applying algorithms for fault detection in-wheel technology (#15). We start with chip/semiconductor and IC related topics (#1 to #5, with the last one a test equipment focus), continue with preassembly (#6) and package/board-related topics (#7 to #9, with the last one a test equipment focus), and finally close with application-related topics (#10 to #15). The newsletter is accessible at Newsletter M24.