DATUM: Determining Authenticity and Trustworthiness of Microelectronics Parts

The aging supply chain management is the available long term protection against IC counterfeiting. For better protection, there is an urgent need for an advanced counterfeit detection system. The industry has made progress in cataloging the features in counterfeit parts and the physical tools to detect them. However, the methods still largely depend on human skills which are time consuming and often not repeatable. Moreover, although there are available techniques to detect changes against a golden sample, there is a need to determine if the changes exclude those made due to the authentic product changes. IAI proposes to develop a cost effective software solution to detect counterfeit parts using machine learning and machine vision algorithms that go beyond traditional approaches. The proposed detection method will analyze images not only from traditional radiography, tomography acoustic images, but also spectroscopy images (e.g., X-ray fluorescence spectrometry, and Fourier transform infrared spectroscopy). The analysis will be tested against the defects catalogue in industry, particularly SAE 6171. The proposed analysis will classify differences between original design and device under test, to guaranty all changes are authorized. The analysis technique will also be adopted to the cases where comparative golden samples are not available.