IEEE IAS Lecture: Learning-Based Approaches for 3D Manipulation of Micro- and Nano-Objects

#protection, #substation, #automation, #relays, #SCADA #WIE

The precise manipulation and assembly of micro- and nano-objects hold immense promise for advancing various industries, from healthcare to nanotechnology. However, the robust and independent manipulation of multiple particles remains a significant challenge due to the global influence of coupled external fields. This presentation will delve into the development of innovative solutions to independently and simultaneously control multiple micro- and nanoparticles in fluid suspensions using shared electric fields. By leveraging adaptive robust motion control and ensemble control systems, we demonstrate how to efficiently manipulate multiple micro- and nanowires using a simple electrode setup in a liquid environment. In addition, I will introduce a learning-based autofocus (AF) and 3D posture estimation scheme that enhances the tracking and manipulation of micro- and nanowires. This approach integrates convolutional neural networks for precise identification of focal distances and inclination angles, enabling the accurate tracking of multiple moving objects in a three-dimensional microfluidic environment. Through transfer learning, we extend the versatility of this system to wires of various materials, achieving high accuracy and efficiency in comparison to traditional methods. These advancements in AF and pose estimation lay the groundwork for automated micro- and nano-object manipulation with wide-reaching applications in material science, nanorobotics, and targeted drug delivery. Finally, I will briefly touch on our ongoing work with Fine-Tuning Hybrid Dynamics, a generalizable physics-informed neural network model, applied to vehicle dynamics but adaptable to other complex dynamic systems. These innovations offer a comprehensive solution to the challenges of autonomous control and manipulation of micro- and nanoparticles, paving the way for next-generation functional nanodevice assembly and advancements in neuromorphic computing.

Kaiyan Yu earned her B.S. degree in Intelligent Science and Technology from Nankai University, China, and a Ph.D. in Mechanical and Aerospace Engineering from Rutgers University, USA. She joined Binghamton University in 2018 as an Assistant Professor in the Department of Mechanical Engineering. Her research focuses on autonomous robotic systems, motion planning, mechatronics, and automation science. In 2022, she received the NSF CAREER Award and currently holds positions as Associate Editor for IEEE Transactions on Automation Science and Engineering, IEEE Robotics and Automation Letters, IFAC Mechatronics, Frontiers in Robotics and AI, the IEEE Robotics and Automation Society Conference Editorial Boards and the ASME Dynamic Systems and Control Division Conference Editorial Boards.