As an alternative to traditional life testing, degradation tests can be effective in assessing product reliability when measurements of degradation leading to failure can be observed. This article ...

Spatial defect patterns generated during integrated circuit (IC) manufacturing processes contain information about potential problems in the processes. The detection of these defect patterns is crucial to improve yield and reliability in IC manufacturing. This paper proposes a multistep defect analysis approach that provides clustering results with different levels of accuracy. A defect denoising step, based on the K th nearest-neighbor noise removal technique, determines the existence of any clustered local defects on a wafer. If local defects exist, the denoising step separates local defects from global defects. A defect clustering step applies a similarity-based clustering technique to group the local defects into clusters according to their spatial locations. A pattern identification step identifies the pattern for each of the local defect clusters (i.e., linear, curvilinear, amorphous, or ring-shaped patterns) via various model selection criteria. Finally, a fine tuning step is applied in order to improve the accuracy of the clustering performance. The fine tuning step is based on model-based clustering with a fixed number of clusters and known patterns for each cluster. The results of both simulated and real wafer map data demonstrate the potential of our approach, both in terms of computational speed and detection accuracy, for analyzing general defect patterns generated during the IC fabrication process.

Accelerated degradation testing (ADT) expedites product degradation by stressing the product beyond its normal use. To extrapolate the product's reliability at use condition, the ADT requires a known functional link relating the harsh testing environment to the usual use environment. Practitioners are often faced with a great challenge to designate an explicit form of the stress-degradation relationship a priori in accelerated degradation models. In this paper, we propose three methods to make direct inference on the lifetime distribution itself without invoking arbitrary assumptions on the degradation model: delta approximation, multiple imputation of failure-times, and the lifetime distribution-based (LDB) method. The methods are easy to implement without computational difficulty, hence they have potential in a wide range of applications for estimating lifetime distributions from ADT data. We applied the methods to two ADT data sets including a real application of commercial organic light-emitting diodes (OLED). The analysis of the examples and simulation results suggests parametric LDB and multiple imputation method as more potential alternatives to traditional failure-time approaches, especially for the case where there is neither enough physical background, nor historical evidence supporting presumed relationships between stress and the parameters of the degradation model.