Physics-Informed Bayesian learning of electrohydrodynamic polymer jet printing dynamics

Calibration of highly dynamic multi-physics manufacturing processes such as electrohydrodynamics-based additive manufacturing (AM) technologies (E-jet printing) is still performed by labor-intensive trial-and-error practices. Such practices have hindered the broad adoption of these technologies, demanding a new paradigm of self-calibrating E-jet printing machines. Here we develop an end-to-end physics-informed Bayesian learning framework (GPJet) which can learn the jet process dynamics with minimum experimental cost. GPJet consists of three modules: the machine vision module, the physics-based modeling module, and the machine learning (ML) module. GPJet was tested on a virtual E-jet printing machine with in-process jet monitoring capabilities. Our results show that the Machine Vision module can extract high-fidelity jet features in real-time from video data using an automated parallelized computer vision workflow. The Machine Vision module, combined with the Physics-based modeling module, can also act as closed-loop sensory feedback to the Machine Learning module of high- and low-fidelity data. This work extends the application of intelligent AM machines to more complex working conditions while reducing cost and increasing computational efficiency.