Printing production quality prediction

A method for providing three:

• dimensional (3D) printing production quality prediction, comprising:
• obtaining a first input associated with a first part of a first plurality of parts during a first 3D print job;
• deriving a first input metric based on the first input;
• mechanically testing at least a portion of the first plurality of parts to determine a ground truth associated with the first input metric;
• and
• training a machine learning function using the input metric and the ground truth resulting from mechanically testing of the portion of the first plurality of parts so that the machine learning function is trained for predicting a likelihood of a second part of a second plurality of 3D parts to satisfy a quality specification by inputting a second input metric to the trained machine learning function during a second print job.

A method for providing three dimensional printing production quality, comprising:

• scanning a region of interest associated with a layer of a plurality of parts during printing of a current print job to obtain an input for the region of interest;
• providing the input for the region of interest to a trained machine learning function that has been trained using scanned input from a corresponding region of interest taken during a number of previous print jobs and results of mechanically testing parts containing that corresponding region of interest from the number of previous print jobs;
• with the trained machine learning function, predicting a likelihood that a part in the current print job containing the region of interest will satisfy a quality specification;
• and
• making adjustments to the current print job based on the predicted likelihood of the part in the current print job containing the region of interest to satisfy the quality specification.