Throughout the years, integrated circuits (ICs) have continuously improved in performance and reduced in cost, allowing them to permeate nearly all human activities, including safety-critical applications such as automotive and biomedical systems. As a result, in the automotive industry, the driving context behind this thesis work, the number of ICs used in vehicles has significantly increased over the past decade. This rise in ICs within automotive systems, however, introduces a significant challenge: the increased likelihood of vehicle failure due to the multiplication of the probabilities of failure of each individual IC. Therefore, to maintain overall product quality, each IC must enhance its own testing quality, with particular focus on latent defects, which are the main contributors to test escapes today.
This doctoral thesis aims to tackle two primary challenges in testing analog and mixed-signal (AMS) ICs in the automotive industry: 1) improving the testing quality by detecting latent defects and 2) reducing the test development time. To address the first challenge, novel testing strategies and design-for-test (DfT) methodologies are explored using a defect-oriented framework. A model for latent defects, specifically for pinholes, is developed and validated experimentally to facilitate simulation-based frameworks for defect-oriented AMS test approaches. The second challenge is addressed through a structured methodology for the test development of analog IP blocks, which allows test reuse and high defect coverage.
This thesis thus provides significant advancements in ensuring the testing quality and reliability of AMS ICs required in automotive applications, while enabling reduction in test development time for ICs where IP reuse is possible.
30/8/2024 17:00 - 19:00
ESAT Aula R
