Topic of smart manufacturing has drawn significant attention in recent years due to its importance in Industry 4.0, its influence to mechatronics industry and its alignment with strategic development in smart machines. To be able to manufacture smart machines, possession of key technologies of smart sensors, smart actuators and smart controllers, especially smart precision sensors, is of major importance for. Starting from reviewing magnetic recording technologies from the speaker’s previous industry R&D with IBM and Hitachi GST Hard Disk Drives, the focus of this workshop will be placed on engineering developments and transformation of magnetic data storage technologies into smart magnetic sensor system for position sensing and control applications in precision machines, overing R&D integration topics of mechatronics, precision mechane design, magnetic materials and manufacturing for the magnetic precision position sensing system.

Torque-sensing of a robotic joint is an important component of future robotic systems that need to have close physical interaction with humans or the environment. Despite decades of studies on the development of different torque sensing technologies, the design of accurate and reliable torque sensors still remains challenging for the majority of the robotics community. This talk will study and evaluate the existing torque sensing methods. Various aspects such as resolution, nonaxial moments load, crosstalk, torque ripple rejection, bandwidth, noise/residual offset level, and stiffness will be discussed. A new method to provide cost-effective torque sensing will be provided toward the realization of higher fidelity joint torque sensing performance.

Estimations of electrical conductivities for biological objects are important to differentiate tissue types and identify altered physiological or pathological conditions. In this talk, the new modeling methods and system development for estimating electrical conductivities by non-contact and contact approaches will be introduced. Two new methods for modeling magnetic/eddy-current and electric fields utilized to develop a bio-magnetic/eddy-current (Bio-M/EC) sensor and an electrical impedance (EI) sensing system for biological object detection will be discussed. With prototypes of Bio-M/EC and EI sensing system, the implementation and measurement procedures for abnormal biological object detection have been illustrated experimentally.

Identification of the inverse dynamics model is essential in precise motion control. We propose a data-driven method for constructing an FIR inverse model. This method applies a generic iterative learning control scheme to track a smoothened impulse signal. Once the tracking converges, the FIR inverse model is concurrently obtained. The constructed inverse model is then used in various control schemes such as feedforward filtering, repetitive control, and inversion-based learning control. The proposed approach has been shown effective on a permanent magnet linear motor, an active magnetic bearing spindle, a galvanometer scanner, and a hydrostatic transmission network