Speaker: Xiao Ding

Topic: Multi-Agent-Based Debate-Driven Decision-Making Technology with Large Language Model

Abstract : Artificial intelligence has progressed from the era of cognitive intelligence to the era of decision intelligence. Human-AI collaborative intelligent decision-making holds significant potential for exploration in both scientific research and practical applications. While contemporary large language models (LLMs) demonstrate remarkable general intelligence and generalization capabilities, their application to decision-making tasks involving complex problems remains a substantial challenge. To address this, this report will present our team's research on multi-agent-based LLM decision-making technology. This includes introducing the concept and principles of the debate-based decision-making method, outlining the key scientific challenges associated with it, presenting our proposed technical solutions, and discussing the future applications of this method within specific vertical domains.

Short Bio: Ding Xiao, Professor and PhD Supervisor at Harbin Institute of Technology (HIT), serves as Deputy Director of the Research Center for Social Computing and Interactive Robotics. His primary research focuses on Artificial Intelligence, Natural Language Processing, Eventic Graphs, and Causal Reasoning. He has published over 90 papers in top-tier international journals and conferences in the AI field, including TKDE, NeurIPS, and ACL. He has led multiple provincial/ministerial-level and national projects, such as those under the National Key R&D Program and the National Natural Science Foundation of China (NSFC). His notable awards and recognitions include: First Prize, Qian Weichang Youth Innovation Award (Chinese Information Processing Science and Technology Award); First Prize and Second Prize, Heilongjiang Provincial Science and Technology Progress Award; First Prize, Wu Wenjun Artificial Intelligence Science and Technology Progress Award; Second Prize, National Teaching Achievement Award; Heilongjiang Computer Society Young Scientist Award; Outstanding Paper Award, ACL 2024; Outstanding Paper Award, NLPCC 2024; First Place, SemEval 2020 International Semantic Evaluation; Named to the AI 2000 Most Influential Scholar in Artificial Intelligence (Global, 2022); Huawei Cloud AI Master Teacher Award. He holds key positions including: Secretary-General, Social Media Processing Committee, Chinese Information Processing Society of China; Deputy Director, Heilongjiang Provincial Key Laboratory of Chinese Information Processing.

Speaker: Zebin Zeng

Topic: "Brain-Body Synergy" Evolution: Research Application and Ecological Practice of Embodied Intelligence Architecture for Humanoid Robots

Abstract : This report focuses on the technology and industrial development of humanoid robots, providing a detailed introduction to the core technical architecture of humanoid robots at the levels of the brain, cerebellum, and body. In the field of scientific research, we closely collaborate with various universities and research institutions to jointly tackle key areas such as motion control and embodied operation. The report will comprehensively showcase the technological breakthroughs and industrialization achievements of KUAVO robots in the direction of embodied intelligence over the past year. The ultimate goal is to work hand in hand with partners from all sectors to jointly build a data ecosystem for the humanoid robot industry, achieving collaborative innovation and sustainable development in the industry.

Short Bio: Zeng Zebin, the head of the humanoid project team at Lepin (Shenzhen) Robotics Technology Co., Ltd., is mainly responsible for the application development of robots. His research directions include: motion control, bipedal robot gait algorithms, and multi-scenario navigation systems, etc. He has led the completion of multiple humanoid robot competition organizations and has collaborated with many universities and research institutions such as Shanghai Jiao Tong University, Beijing Institute of General Artificial Intelligence, Shandong University, and Tongji University to conduct joint research projects and output scientific research results in the field of robotics.

Speaker: Wenzheng Chi

Topic: Enhancing Adaptive Interaction in Complex Environments: Embodied Intelligence for Mobile Robot Perception and Social Collaboration

Abstract : This study investigates intelligent perception and social interaction capabilities for mobile robots operating in complex environments, driven by embodied intelligence. Employing an embodied cognitive framework, the research establishes a closed-loop process enabling robots to progress from environmental understanding to human-robot collaboration. The proposed approach aims to provide autonomous and humanized solutions for critical application scenarios such as service robotics and smart cities. Ultimately, this work seeks to advance the level of adaptive interaction and intelligent decision-making for robots functioning in open-world settings. The framework integrates environmental sensing, cognitive processing, and interactive behaviors to foster more natural and effective cooperation between robots and humans within dynamic real-world contexts.

Short Bio: Professor Chi is an IEEE Senior Member and recipient of the Jiangsu Provincial ‘333 High-Level Talent Training Project’ and the Jiangsu Association for Science and Technology ‘Young Talent Lift Project’. She earned her Ph.D. Degree from The Chinese University of Hong Kong, which included a research visit to the University of Tokyo, Japan. Since joining Soochow University in 2018, she has led research in Embodied AI and Robotics, specializing in robot motion planning and human-robot interaction (HRI). Professor Chi has secured significant funding as Principal Investigator, including projects for National Key Research and Development Program of China and the National Natural Science Foundation of China. She has authored over 60 SCI/EI papers (including 2 ESI Highly Cited and 10 Top-tier journal papers) and holds 20+ patents. Her contributions have been recognized with awards such as the Second Prize in the Science and Technology Award (Jiangsu Association of Automation, 2021), and the Best Student Conference Paper Award at IEEE RCAR (2016, 2021).

Speaker: Zhiwei Cui

Topic: Development and application of embodied intelligent robots

Abstract : This report primarily introduces the concept and definition of embodied intelligence and examines its research significance from several perspectives, including the historical development of robotics and artificial intelligence, the growing market demand for embodied intelligent robots, and their role as engines of new quality productivity. Furthermore, drawing on the latest research findings from the embodied intelligence robot research team, the report elaborates on the application scenarios of embodied intelligent robots and summarizes the key challenges and opportunities associated with the development of embodied intelligence technologies, along with future outlooks.

Short Bio: Cui Zhiwei, School of Artificial Intelligence, Shandong University, Professor, Doctoral Supervisor. He received the Ph.D. degree from Tsinghua University and conducted postdoctoral research at The Chinese University of Hong Kong. His research interests include medical robots, embodied intelligent robots, robot intelligent control and interaction, and the structural design and analysis of robotic systems.