Vision based Iterative Learning Control for a roll to roll micro/nano-manufacturing system

Fabrication of nano/micro-scale functional devices oftentimes involves multiple steps. In the context of a continuous or semi-continuous manufacturing process, each fabrication step is performed successively in multiple localized zones. As the substrate or the web traverses downstream in the process flow, proper registration of the pre-existing features is necessary prior to entering the next fabrication zone in order to accurately complement previous manufacturing steps. By performing a direct observation of the pre-existing feature using machine vision, the uncertainty of the feature location can be circumvented. In this paper, a two-layer feedback architecture with a vision sensor in the outer loops is used to visually servo the pre-existing feature on the web. Additionally, the feedback controller is augmented with Norm Optimal Iterative Learning Control (NOILC) to improve the position tracking and tension regulation of the web. Simulation and experimental results show the advantages of implementing NOILC in the R2R system.