Advanced sensing presents the prerequisite for realizing intelligent manufacturing. Sensors monitor production operations in real-time, often in harsh environments, provide input for diagnosing the root cause of quality degradation and fault progression such that subsequent corrective measures can be formulated and executed online to control a machine’s deviation from its optimal state. With the increasing convergence among measurement science, information technology, wireless communication, and system miniaturization, sensing has continually expanded the contribution of mechatronics to intelligent manufacturing, enabling functionalities that were not feasible before in terms of in-situ state monitoring and process control. New sensors not only acquire higher resolution data at faster rates but also provide local computing resources for autonomously analyzing the acquired data for intelligent decision support.
This talk presents research on advanced sensing for improved observability in manufacturing process monitoring, using polymer injection molding and sheet metal micro rolling as two examples. The design, characterization, and realization of multivariate sensing and acoustic-based wireless data transmission techniques in RF-shielded environment are first introduced. Next, computational methods for solving an ill-posed problem in data reconstruction are described. The talk highlights the significance of advanced sensing and data analytics for advancing the science base and state-of-the-technology to fully realize the potential of intelligent manufacturing.
