Machine Learning applied to Measurement Assessment

Areté’s SQUIRE (Signal Quality Utility for Identification and Removal of Errors) leverages the latest in machine and deep learning to automatically detect, classify, and localize sources of contamination in radar returns. Currently, test engineers at NRTF use the SABER software tool to perform analyses to manually identify and localize contamination in radar data that needs to be removed before further processing. SQUIRE will automatically identify contamination in radar imagery to support test engineers and significantly shorten this time consuming process. Expanding our successful Phase I effort, in Phase II a prototype SQUIRE implementation will be integrated with SABER. SQUIRE will be extended to additional radar bands utilized by NRTF and include additional contamination sources. SQUIRE will be designed to provide an easy mechanism for operator feedback to enable straightforward updates to its machine learning networks. SQUIRE integration with SABER allows model updates to be simple and seamless. The envisioned drop-in configuration file means software recompilation and full code stream delivery will not be required, thus making updates easy and straightforward to the end user. A successful SQUIRE integration will benefit other radar test facilities to help elevate operator overload.