The industry sponsor for my project was a large, well-known oil services company, who wanted to create an automated system of grading the wear on drill bits used for drilling wells. Our solution was developed into a working web application prototype, containing multiple subsystems. These included methods to extract objects using a custom-trained machine learning model in TensorFlow, and multiple novel image processing methods for detecting wear and damage to drill bits. Despite a limited dataset, our final prototype determined damage and wear with a 86% accuracy when compared to hand-labeled damage.