Pannier Research Lab

The standard approach in additive manufacturing (or 3D printing) is to "slice" a long list of machine instructions (thousands to millions of lines of G-code) assuming the process fits a model, upload the sliced instructions to the machine, and then to press "Start". Unfortunately, today's additive manufacturing machines lack the feedback signals required to determine whether the part under construction is being built correctly, that is, within the design tolerance. The build process commonly suffers from disturbances (changing dynamics) so that the sliced machine instructions, even if executed accurately, will not result in a printed part within design tolerance. Even worse, the act of slicing  all the instructions upfront (feedforward only control) presupposes a process model that may well be out-of-calibration. In this lab, we explore control solutions that incorporate feedback and estimation into the build process to compensate for disturbances or changing  dynamics and thereby achieve improved design tolerances.