Spatial virtual metrology prediction using multimodal machine data

Abstract

Described are systems and methods for predicting an overlay of a patterned wafer. Methods can include receiving a machine data set comprising parameters at a plurality of positions of a wafer, generating an integrated data set by aggregating the parameters and mapping the aggregated parameters to an equispatial grid, generating, using a trained machine learning (ML) model, a predicted overlay for the patterned wafer based at least in part on processing the integrated data set.