Machine learning assessment of autonomous drill bit hard face quality

A system comprising:

• a data repository for storing a three-dimensional scan of a coating that coats an object, wherein the coating and the object, in combination, have a shape;
• and
• a processor in communication with the data repository and programmed to:
• flatten the three-dimensional scan to generate flattened data, wherein:
• flattening comprises conforming scan data in the three-dimensional scan to zone data for a plurality of zones of a layout,
• the plurality of zones represents a plurality of sections of the shape cut along a plurality of boundaries in the shape, and
• each of the plurality of zones represents a corresponding thickness of the coating in a corresponding section of the plurality of sections;
• vectorize the flattened data into a vector data structure;
• and
• execute a machine learning model, wherein:
• the machine learning model takes, as input, the vector data structure and produces, as output, a prediction whether one or more zones of the plurality of zones are out of an engineering tolerance for the coating.

A method comprising:

• receiving a three-dimensional scan of a coating that coats an object, wherein:
• the three-dimensional scan comprises a point cloud,
• each point in the point cloud is expressed in four coordinates, and
• the four coordinates are expressed as three dimensional coordinates plus a fourth coordinate that represents a height of the coating above a surface of the object;
• vectorizing the point cloud into a vector data structure;
• and
• applying a neural network to the vector data structure, wherein an output of the neural network comprises an output vector that includes a prediction of whether at least one point in the point cloud is within an engineering tolerance defined for the coating.