Image-Based Machine Learning for Component Identification for Remanufacturing

This project proposes to develop an automated image-based part type identification in the production environment. The goal is to implement the process at CoreCentric’s remanufacturing floor and demonstrate a 50% increase in processing speed and a 20% yield increase. Additionally, this project aims to remove subjectivity from the process of identification and inspection.

Upon completion, this project expects to reduce primary feedstock by 67,000 metric tons (MMT) of metals, polymers, and e-waste, reduce energy consumption by 13 PJ, and decrease emissions by 1.5 MMCO2e.