おしらせ (News!!)

国際ロボット展(iREX)とUnitree Conference 2025で発表します

12月4日
「複数感覚情報のマルチモーダルAIが切り拓く！器用さを備えるAIロボット」https://irex.nikkan.co.jp/webinar/detail/1607

12月16日「ロボットシステムのための生体模倣AIに向けて」

https://techshare.co.jp/riw2025/unitree-developer-conference2025/?gad_source=1&gad_campaignid=23202794308&gclid=Cj0KCQiAosrJBhD0ARIsAHebCNo5F1aY8YryehQ4rhl50gsV663JyWUUlbE-n0ctsfF5klLyJOnL0NgaAlkjEALw_wcB

I gave Keynote Talk at CACRE2025, held at Wuxi, China from July 16 to 19, 2025.
https://www.cacre.org/index.html
This year was 10th Edition, and sponsored by IEEE RAS, The Hong Kong University of Science and Technology (HKUST) and Hong Kong Society of Mechanical Engineers(HKSME).
It was great opportunity to meet important people from Hong Kong Society of Mechanical Engineers and other keynote speakers.

産経新聞社「第38回 独創性を拓く 先端技術大賞」において、ロボティクス専攻 博士後期課程3年の杉山拓さんが学生部門「ニッポン放送賞」を受賞することが発表されました。
本賞は優れた研究成果を上げた理工系学生や企業・研究機関などの若手研究者、技術者の論文を表彰するものです。
7月14日（月）に授賞式が執り行われる予定です。

https://www.sankei-award.jp/sentan/jusyou/

https://www.eng.tohoku.ac.jp/news/detail-,-id,3250.html

ブリヂストンと学習と情報処理に関する共同研究がスタートしました（2023-2024）

1. Learn to Navigate in Dynamic Environments with Normalized LiDAR Scans, IEEE Int. Conf. on Robotics and Automation, May 2024, W. Zhu, M. Hayashibe (ICRA2024)

3名の博士課程学生がJSPS 特別研究員に採択されました

杉山拓(DC2) ,Yuchen Wang (DC2) 古関駿介 (DC1)

1. Motor Synergy Development in Symmetric Gait of Whole-body Locomotion Learning, IEEE Int. Conf. on Robotics and Automation, May 2023, G. Li, A. d’Avella, M. Hayashibe (ICRA2023Video)
2. Learnable Tegotae-based Feedback in CPGs with Sparse Observation Produces Efficient and Adaptive Locomotion, IEEE Int. Conf. on Robotics and Automation, May 2023, C. Herneth, M. Hayashibe, D. Owaki (ICRA2023)
3. Morphological Characteristics That Enable Stable and Efficient Walking in Hexapod Robot Driven by Reflex-based Intra-limb Coordination, IEEE Int. Conf. on Robotics and Automation, May 2023, W. Sato, J. Nishii, M. Hayashibe, D. Owaki (ICRA2023)

2 papers have been accepted for ICRA2021!

1st one is about synergy-redundancy coupling, and its usage for quantification of redundancy considering given task and dynamic feasibility through deep learning.
2nd one is about softrobotics swim simulation and its deep learning framework.

1. Quantification of Joint Redundancy considering Dynamic Feasibility using Deep Reinforcement Learning, IEEE Int. Conf. on Robotics and Automation, June 2021, J. Chai, M. Hayashibe (ICRA2021)
2. Deep Reinforcement Learning Framework for Underwater Locomotion of Soft Robot, IEEE Int. Conf. on Robotics and Automation, June 2021, L. Guanda, J. Shintake, M. Hayashibe (ICRA2021)

3 papers have been accepted for ICRA2020 Paris!!

1. Simultaneous On-line Motion Discrimination and Evaluation of Whole-body Exercise by Synergy Probes for Home Rehabilitation, IEEE Int. Conf. on Robotics and Automation, (2020), (accepted), Felipe M. Ramos, M. Hayashibe, (ICRA2020 Paris)  This paper is important work for On-line Synergy Tracking which allows simultaneous motion discrimination and motion evaluation with home setting!!
2. Motor Synergy Development in High-performing Deep Reinforcement Learning algorithms, IEEE Robotics and Automation Letters, (2020),(accepted), Jiazheng Chai, M. Hayashibe, (ICRA2020 Paris) This paper is important work for revealing what is going on under deep reinforcement learning in motion learning. Spatio-temporal synergy is emerged in the course of learning without specifying any synergy usage!!
3. Discovering Interpretable Dynamics by Sparsity Promotion on Energy and the Lagrangian, IEEE Robotics and Automation Letters, (2020), (accepted), Hoang K. Chu, M. Hayashibe, (ICRA2020 Paris) This paper is important work for discovering analytical dynamics with experimental observation. Sparse Analytical models for unknown dynamics could be obtained from observation of mass motion with white-box approach!!

Identification of Time-Varying and Time-Scalable Synergies From Continuous Electromyographic Patterns, IEEE Robotics and Automation Letters, (2019), Felipe M. Ramos, Andrea d’Avella, Mitsuhiro Hayashibe, (accepted both for RAL journal and IROS2019 Macau)

Existing Time-Varying synergy may detect different speed motions as different synergy combination as time component is different between slow and fast movement. However, in neuroscience perspective, it should come from common synergy just with different speed control. The proposed method allows to make new analysis for auto-identify the common synergy over different speed motions during the synergy extraction.