The 7th ZJU Graduate International Summer School on“Distributed Control, Optimization and Learning”Was Successfully Held

From August 2 to 13, 2021, the ZJU Graduate International Summer School on Distributed Control, Optimization and Learning, organized by the Network Sensing and Control Research Group, College of Control Science and Engineering, Zhejiang University, was successfully held in the Yuquan Campus. The summer school consists of 9 guest lectures and 7 seminars delivered by leading research scholars from Zhejiang University, Chinese University of Hong Kong, Tufts University, Rice University, Arizona State University, Purdue University, Carleton University, Nanyang Technological University, Pennsylvania State University, Nankai University and Alibaba Damo Academy, respectively. This event attracts more than 200 graduate students from the worldwide. The summer school is designed with two modules. The first module is dedicated to the teaching of basic knowledge in the field of distributed control and optimization, while the second module is for bringing students frontier advanced topics related to distributed optimization. This summer school adopted the combination of online and offline teaching, with offline teaching mainly for students of Zhejiang University and online teaching for students from other universities.Offline participationOnline participationThe opening ceremony on August 2 marked the start of the event. Professor Jinming Xu from the College of Control Science and Engineering of Zhejiang University presided over the opening ceremony and taught the first class, introducing some basic knowledge of convex optimization and explaining his profound understanding of optimization theory. Professor Xu's vivid explanation stimulated students' interest in distributed control and optimization, and the following courses and activities were carried out very smoothly. In the first week of the basic courses, the professors gave a detailed explanation of convex optimization, graph theory and consistency, distributed stochastic optimization, large-scale machine learning, distributed control and other related basic knowledge.In the second week of the lecture, Professor Gesualdo Scutari from Purdue University introduced some new ideas and novel analyses of distributed optimization from the statistical perspective, and pointed out the underlying reasons for the failure of distributed learning in high-dimensional scenarios and the mismatch between experimental results and theoretical analysis. Professor Usman Khan from Tufts University introduced a novel algorithm framework that enables distributed learning on non-convex optimization problems, and demonstrated the reliability of his proposed method with provable theoretical results and numerical experiments with real data. Professor Cesar Uribe from Rice University explained the optimal complexity of distributed optimization algorithms, introduced a scalable algorithm that can achieve the same convergence rate as centralized counterparts, and shared some interesting application examples of distributed optimization and learning. Professor Lihua Xie from Nanyang Technological University introduced a complete set of smart sensing and localization technologies for IoT and unmanned systems, such as WiFi-based indoor positioning and human activity recognition, UWB-based positioning, and Visual-Inertial-Distance sensor fusion for positioning and mapping, and their wide interesting applications in various domains. Last but not the least, Professor Hoi To Wai from Chinese University of Hong Kong introduced the recent techniques for optimizing distributed nonconvex models  that process batch/streaming data, and explained how to balance communication and computation complexity to design efficient algorithms. Professor Shichao Liu from Carleton University introduced a co-design scheme that properly combines distributed control and event-triggered scheduling so as to meet the challenges of load frequency control in multi-area smart grid. These informative and wonderful lectures and seminars left a deep impression on the students, which would be beneficial to their own research!In addition to the impressive online lectures, the summer school also provided lab visits and academic seminars for the students. During the lab visits, Professor Xu led the students to visit the smart micro-grid laboratory, industrial control safety laboratory (involving PLC, mechanical arm safety) and anti-aerial-robotic systems laboratory, which helped students have a deep understanding of intelligent micro-grid technology, industrial control system attack and defense technology, and anti-anti-aerial-robotic technology based on multi-sensor fusion, and have a further understanding of the engineering application of distributed control and optimization.Students of Zhejiang University participated in lab visitsThis summer school is an important measure for the School of Control Science and Engineering of Zhejiang University to build a world-class discipline and promote the level of internationalization. It also promoted the academic exchange between students and broadened their horizons. This summer school ended successfully!Reporter: YAN ChangzhiEditor: HE Yushan

Dialogue@ZJU: With international faculty members and students on the trip to Anji

A glimpse into ZJU’s gymnasium for Asian Games 2022

Dialogue@ZJU: Congratulations! Class of 2021

“Fast and Furious”: The 2021 Int’l Concrete Dragon Boat Competition

ZJU among the top 50 universities in the world

Celebrating 124 years of excellence

Dr. GAO Fei received Honorable Mention in 2020 IEEE-TRO King-Sun Fu Memorial Best Paper Award as first author

Recently, Dr. GAO Fei, Associate Professor of the Field Autonomous System & Computing Lab (Fast Lab), College of Control Science and Engineering, received Honorable Mention in the 2020 IEEE Transactions on Robotics (IEEE-TRO) King-Sun Fu Memorial Best Paper Award for his paper Teach-Repeat-Replan: A Complete and Robust System for Aggressive Flight in Complex Environments as the first and corresponding author. This is the first time that the colleges and universities of the Chinese Mainland won the honor as the first completion unit. It is also the first time that IEEE-TRO awarded this honor to the papers published in IEEE-TRO on aerial robotics.There is a massive market for consumer drones nowadays. However, most of the operators of consumer drones are not professional pilots and would struggle to generate their ideal trajectory for a long time. In some scenarios, such as drone racing or aerial filming, it is impossible for a beginner-level pilot to control the drone to finish the race safely or take an aerial video smoothly without months of training. In a drone racing competition, each quadrotor is controlled by a human pilot to fly through several gates towards the terminal as quickly as possible. In the racing flight, collisions must be avoided to ensure safety, while the flight aggressiveness is expected to be extremely high. However, it is hard for a human pilot to master the skill of balancing speed and safety. As opposed to drone racing, aerial filming/videography does not prefer high speed, but good motion smoothness, because gentle transitions are typically good for generating aesthetical videos.Based on the above observations, this paper presents a complete solution towards robust aerial autonomy, enabling a drone to accomplish a complicated task with professional performance under merely rough human operations. The human operator may provide an arbitrarily slow or jerky trajectory with an expected topological structure. The system then autonomously converts this poor teaching trajectory to a topologically equivalent and local optimal one. The aggressiveness of the generated repeating motions is tunable, which can meet speed requirements ranging from drone racing to aerial filming. Moreover, during the repeating flight, the system locally observes environmental changes and replans safe trajectories to avoid moving obstacles. The proposed system extends the classical robotics teach-and-repeat framework and is named as Teach-Repeat-Replan.Before the announcement of this news, the work was ranked among the top 50 hot articles in the IEEE-TRO database (ranked 14th), and was widely concerned and highly praised by eminent scholars at home and abroad. Kevin Lynch, chief editor of IEEE Transactions on Robotics, Professor of Northwestern University, author of IEEE fellow, Modern Robotics and other reputable robotics textbooks, when awarding the King-Sun Fu Best Paper Award Honorable Mention, commented, “The T-RO editorial board was impressed by the theoretical and experimental contributions of your work.”IEEE-TRO is one of IEEE’s top journals in the field of robotics, including all aspects of research in this sector. IEEE-TRO has very high requirements for the quality of papers, which demands outstanding theoretical and practical contributions to the field of robotics that can promote the development of robotics. The “King-Sun Fu Best Paper Award Honorable Mention” is established to recognize the best work in the papers published by IEEE-TRO every year. Therefore, it is a high honor for all robot researchers to publish papers in IEEE-TRO and win the award at the same time.Related Links:About the Paper: https://ieeexplore.ieee.org/document/9102390Open Source of the Work: https://github.com/HKUST-Aerial-Robotics/Teach-Repeat-ReplanVideo: https://www.bilibili.com/video/BV1Fx411o78w?from=search&seid=4539658003138370390The FAST Lab: http://zju-fast.comReprinted from: The Fast LabReporter: GAO FeiEditor: WANG Jing

Congratulations graduates

Dialogue@ZJU：With Prof. LI Lanjuan, a member of the Chinese Academy of Engineering