Skip to main content

RBE Colloquium Series: Dr. Yayun Du | Adaptive Autonomous Cyber-Physical Systems: Design, Perception & Control

Thursday, March 17, 2022
2:00 pm to 3:00 pm


Floor/Room #: 
Room 400

Robotics Engineering Colloquium Series

Special Presentation

Dr. Yayun Du

University of California, Los Angeles

Adaptive Autonomous Cyber-Physical Systems: Design, Perception, & Control


Thursday, March 17th 2022

2:00 PM - 3:00 PM

Unity Hall | Room 400 |


Abstract: Cyber-Physical Systems (CPS) are ubiquitous in engineering, with applications in agriculture, autonomous vehicles, health and medicine, manufacturing, robotics, and aeronautics. Robust autonomous CPS require a seamless integration of design, sensing (perception), control, and coordination. This complexity introduces fundamentally new challenges and requires a dramatic increase – compared with the state of the art – in the number, precision, and sampling rate of modern sensors. As such, CPS are becoming more and more computation- and data-intensive. Moreover, CPS are meant to operate in the physical world, which typically faces continual dynamic changes, feature events, and external conditions that are hard to predict or model, e.g., other CPS and human agents. All these aspects pose open research challenges. My doctoral research on biolocomotion develops the simplest robots to verify scientific concepts, e.g., bacterial “turning-by-buckling” mechanism. Meanwhile, we conducted analytical analyses and augmented the Discrete Elastic Rods (DER) simulation tool to simulate the mechanics of the locomotion in granular media and viscous fluids. DER was originally developed in the computer graphics community to simulate hair in movies (e.g., The Hobbit). My research successfully synergizes computational mechanics and design: robots are developed to validate simulators, and validated simulators serve as data generators for robot design via machine learning (ML). Apart from rapid-prototyping locomotion designs, my research focuses on the design and evolution of functional robots, such as autonomous under-canopy agricultural robots. They interact with the environment for functionality, addressing the challenges faced by traditional CPS. My long-term objective is twofold: (i) Build on physics knowledge and ML to develop individual adaptive autonomous CPS and (ii) Translation to practice. Potential areas of application of these CPS include biomimetic robotics, agricultural robotics, soft wearable devices, reconnaissance devices, autonomous robotic surgery, collaborative robotics, underwater preservation of fragile organisms and samples, autonomous driving, and intelligent transportation systems.

Bio: Yayun Du is a Ph.D. candidate in Mechanical & Aerospace Engineering at the University of California, Los Angeles (UCLA), with an expected graduation date of June 2022. She obtained a B.S. in Automotive Engineering from the Harbin Institute of Technology in 2016 as one of the university’s “Top Ten Students.” She earned her M.S. in Systems and Control at UCLA in 2018. In 2021, she was also awarded “MIT Civil and Environmental Engineering (CEE) Rising Star”. Her research interests lie at the interface of robotics, control, mechanics, and artificial intelligence. The core contributions of her doctoral research are: (1) developing and modeling bio-inspired robotic locomotion (e.g., untethered flagellar robots in granular media and viscous fluids, starfishinspired robots), (2) developing autonomous low-cost agricultural robots for under-canopy weed control and phenotyping, and (3) controlling robotic arms to achieve tasks (e.g., painting walls) autonomously.