Join the CALCE/SMTA Counterfeit Parts and Materials Symposium 2022 for this and other informative presentations
Christopher P. Walker
Abstract: This work aims at layer-to-layer evaluations of a printed circuit board (PCB) using differential imagery analysis to target board-specific changes. Utilizing Non-destructive techniques for hardware assurance coupled with internally developed software, the motivation of this approach is to enable high throughput comparisons of “Golden Copy” to an evaluation article for supply chain validation and verification. The intent is to detect non-conformances on the outer and inner layers of a populated printed circuit board (PCB) by doing a layer-by-layer comparison and evaluation of the “Golden Copy” against the device under test (DUT). This new data flow is enabled by layer extraction of the PCB, comparison of those layers, and a future tool for Gerber overlay. Additional discussion will center around the varying density of materials and components on the surface of PCB’s which introduce image artifacts whose mitigation requires unique approaches to deal with image errors Because of these artifacts, image processing errors can cause increased human drive time, which can impact the time spent per article when performing validation and verification (V&V) efforts.
Bio: Christopher Walker received a B.S. in Computer Science & Applied Mathematics in 2004, and an M.S. in Computer Science in 2005 from the University of Missouri-Rolla. Chris has spent much of his seventeen-year career at Sandia developing and deploying image processing and machine learning algorithms supporting a wide of a variety domains, with the last seven years primarily focused on improving analysis of X-Ray, computed tomography, and scanning electron microscope data used in destructive and non-destructive microcircuit analysis.
Top