Bringing human dexterity to robots by combining human motion and tactile sensation
Researchers develop an adaptive motion system that allows robots to generate human-like movements with minimal data
Keio University Global Research Institute
image:
This image depicts the real-time transfer of a human’s motion to a robotic avatar, enabling the latter to perform a dexterous task.
view moreCredit: Associate Professor Takahiro Nozaki from Keio University, Japan
Accelerating progress in robotic automation promises to revolutionize industries and improve our lives by replacing humans in risky, physically demanding, or repetitive tasks. While existing robots already excel in controlled environments such as assembly lines, the ultimate frontier of automation lies in dynamic environments found in tasks, such as cooking, assisting the elderly, and exploration. To realize this goal, one of the key barriers is making robots capable of adapting to touch. Unlike human hands, which intuitively adjust their grip for objects of unknown weight, friction, or stiffness, most robotic systems lack this crucial form of adaptability.
To transfer sophisticated human dexterity to machines, researchers have developed various motion reproduction systems (MRSs). These are centered around accurately recording human movements and recreating them in robots via teleoperation. However, MRSs tend to encounter problems if the properties of the object being handled change or do not match those of the recorded movement. This limits the versatility of MRSs and, in turn, the applicability of robots in general.
To address this fundamental challenge, a research team from Japan has developed a novel system designed to adaptively model and reproduce complex human motions. The study was led by Master’s student Mr. Akira Takakura from the Graduate School of Science and Technology, Keio University, and co-authored by Associate Professor Takahiro Nozaki, Department of System Design Engineering; Doctoral student Kazuki Yane; Professor Emeritus Shuichi Adachi, also from Keio University; and Assistant Professor Tomoya Kitamura from Tokyo University of Science, Japan. Their paper was published in IEEE Transactions on Industrial Electronics, a world-leading international academic journal in this field, on December 30, 2025.
The team’s core breakthrough was moving past linear modeling strategies and instead using Gaussian process regression (GPR). This is a regression technique that can accurately map complex nonlinear relationships, even with a small amount of training data. By recording human grasping motions for multiple objects, the GPR model was trained to identify the relationship between the object’s ‘environmental stiffness’ and the necessary position and force commands issued by the human. In turn, this process effectively reveals the human’s underlying motion intention, or ‘human stiffness’—allowing the robot to generate appropriate motion for objects it has never encountered. “Developing the ability to manipulate commonplace objects in robots is essential for enabling them to interact with objects in daily life and respond appropriately to the forces they encounter,” explains Dr. Nozaki.
To validate their approach, the researchers tested it against conventional MRSs, linear interpolation, and a typical imitation learning model. The proposed GPR system demonstrated significantly enhanced performance in reproducing accurate motion commands for both interpolation and extrapolation. For interpolation, which involves handling objects with a stiffness that falls within the limits of the training set, it reduced the average root-mean-square error (RMSE) by at least 40% for position and 34% for force. Meanwhile, for extrapolation of objects harder or softer than those in the training set, the results were equally robust, exhibiting a 74% reduction in position RMSE. Most importantly, the proposed method based on GPR markedly outperformed all other methods.
By accurately modeling human–object interactions with minimal training data, this new take on MRSs will help generate dexterous motion commands for a wide range of objects. This ability to capture and recreate complex human skills will ultimately enable robots to move beyond rigid contexts and toward providing more sophisticated services. “Since this technology works with a small amount of data and lowers the cost of machine learning, it has potential applications across a wide range of industries, including life-support robots, which must adapt their movements to different targets each time, and it can lower the bar for companies that have been unable to adopt machine learning due to the need for large amounts of training data,” comments Mr. Takakura.
Worth noting, this research group at Keio University has been actively engaged in research concerning the transmission, preservation, and reproduction of force-tactile feedback. Their previous efforts in this field have covered a wide range of topics, such as the reduction of data traffic, motion modeling, and haptic transplant technology. Their groundbreaking work on sensitive robotic arms and ‘avatar’ robots has been widely recognized by electronics research institutions like the IEEE, as well as by organizations such as the Government of Japan and Forbes.
***
Reference
DOI: 10.1109/TIE.2025.3626633
About Keio University Global Research Institute (KGRI), Japan
The Keio University Global Research Institute (KGRI) was established in November 2016 as a research organization to bridge faculties and graduate schools across the university. KGRI aims to promote interdisciplinary and international collaborative research that goes beyond the boundaries of singular academic disciplines and international borders. It also aims to share research outcomes both in Japan and worldwide, further promoting engagement in joint research.
To achieve this goal, KGRI has set up more than 40 centers and projects funded by external sources or through internal grants, covering a wide range of research topics from basic research to addressing social challenges facing the world. In 2022, Keio University set its goal of becoming a "Research university that forges the common sense of the future".
Website: https://www.kgri.keio.ac.jp/en/index.html
About Associate Professor Takahiro Nozaki from Keio University
Dr. Takahiro Nozaki received his B.E., M.E., and Ph.D. from Keio University, Yokohama, Japan, in 2010, 2012, and 2014, respectively. In 2014, he joined Yokohama National University, Yokohama, Japan, as a Research Associate. In 2015, he joined Keio University, where he is currently an Associate Professor. He was also a Visiting Scientist with the Massachusetts Institute of Technology, Cambridge, USA, from 2019 to 2021. He was one of the winners of the IEEE Industrial Electronics Society Under-35 Innovators Contest in 2019.
https://k-ris.keio.ac.jp/html/100011714_en.html
About Mr. Akira Takakura from Keio University
Mr. Akira Takakura received a B.E. degree in System Design Engineering from Keio University, Yokohama, Japan, in 2024. He is currently working toward an M.E. degree. His research interests include adaptive control, system identification, robotics, and haptics.
Funding information
This work was supported by JSPS (Grant No. 16H06079) and NEDO (Project No. P15009, “Development of Core Technologies for Next-Generation AI and Robotics”).
Journal
IEEE Transactions on Industrial Electronics
Method of Research
Experimental study
Subject of Research
Not applicable
Article Title
Motion Reproduction System for Environmental Impedance Variation via Data-Driven Identification of Human Stiffness
From lab to rubble: can this lightning-fast origami robot be the future of search and rescue?
Scientists develop a powerful modular soft origami actuator for building small-scale soft robots with exceptional shape reconfiguration and locomotion abilities
Journal Center of Harbin Institute of Technology
Imagine soft robots scurrying through rubble with the agility of geckos or leaping over obstacles with the dynamism of frogs—capabilities that would prove invaluable in post-disaster rescue and exploration missions. In recent years, soft robotics has emerged as a prominent focal point of academic research, owing to its intrinsic compliance, inherent operational safety, and superior environmental adaptability. Yet, endowing soft robots with biological-grade agility and speed has long remained a pivotal challenge that researchers strive to address.
Currently, most soft robots face a "performance shackle" in practical applications: they are often slow and weak in mobility. This limitation stems from a trade-off in soft actuators, where it is difficult to achieve high-speed response, large strain and force output simultaneously. Consequently, soft robots have long been stuck in a "slow-motion" phase, unable to perform high-mobility tasks in the real word.
To overcome the inherent "sluggishness" of traditional soft robots and enable high-speed, agile movements on par with their natural counterparts, a research team from Tongji University has carried out in-depth research—findings recently published in the journal SmartBot. Inspired by the synergistic "skeletal-muscle" mechanism of vertebrates, the team proposed a "rigid-flexible" solution: merging the geometric topology of origami with high-performance electrohydraulic actuators. This novel Electrohydraulic Origami (EHO) actuator uses a hexagonal rigid frame to act as a "skeleton" and electrohydraulic units to serve as "muscles". By applying high voltage to the flexible electrodes, the resulting electrostatic force squeezes the liquid dielectric, thereby driving the rotation of the origami joints. This design cleverly leverages the amplification and transmission principles of both structure and mechanics, transforming minute electrohydraulic zipping motions into massive, high-speed and multimodal structural deformations of the origami framework.
By integrating soft electrohydraulic actuation with origami structures, the EHO actuators unlock entirely new actuation modes while delivering remarkable dynamic performance. Experimental results indicate that an EHO actuator with a 10-cm arm length can achieve a staggering axial strain of up to 3300% and a peak strain rate exceeding 23500 %/s—outperforming most existing soft actuators. Meanwhile, a smaller 2.5-cm-arm version can perform vertical jumps 8.5 times its own height without any external energy storage components. Furthermore, even under a 100 g load—equivalent to 12.5 times its own weight—the actuator maintained an impressive 483.3% actuation strain, demonstrating exceptional load-bearing capacity. Various robotic prototypes powered by EHO actuators demonstrate exceptional locomotion capabilities. Notably, the tethered crawling soft robot can achieve an average linear speed of 37.55 cm/s (approximately 9.39 body lengths per second) and a rapid turning velocity of 211.5°/s. Furthermore, the wireless version, equipped with a miniature high-voltage control system, delivers average speeds that outpace most previously reported untethered crawling soft robots driven by electroactive soft actuators. Moreover, the robot can crawl steadily on a 15.6° incline and perform agile maneuvers along S-shaped paths. These capabilities mark a significant milestone in transitioning electrohydraulic robots from laboratory prototypes toward untethered, practical applications like field exploration, search-and-rescue, and human-machine interactions, etc.
About Tongji University
Tongji University is a prestigious comprehensive research university located in Shanghai, China. As a member of the "Double First-Class" initiative, it excels in engineering, architecture, urban planning, and environmental science, with several disciplines ranked among the world’s top. Boasting world-class faculty, cutting-edge laboratories, and extensive global collaborations with leading institutions, Tongji prioritizes innovative research and interdisciplinary education. It nurtures talented professionals dedicated to addressing national and global challenges, making remarkable contributions to urban development, sustainable technology, and social progress.
Website: https://www.tongji.edu.cn/
About Dr. Wenbo Li from Tongji University, China
Dr. Li graduated from Shanghai Jiao Tong University, China in 2019 with a Doctor of Philosophy (PhD) degree in Mechanical Engineering. After completing a two-and-a-half-year postdoctoral research fellowship at the State Key Laboratory of Mechanical System and Vibration, he joined Tongji University as a research professor of the School of Aerospace Engineering and Applied Mechanics in 2022. His research interests include soft robots, smart materials and structures. He has published about 40 peer-reviewed papers in journals including Nature Communications, Science Advances, SmartBot, Advanced Functional Materials, etc. He also holds approximately 20 authorized Chinese patents.
Funding information
This work was supported by National Natural Science Foundation of China (Grants 12472057, 12532002, and 12102398), General Project of the Shanghai Natural Science Foundation (Grant 24ZR1468800), Research Project of State Key Laboratory of Mechanical System and Vibration (Grant MSV202407), Fundamental Research Funds for the Central Universities, and Shanghai Gaofeng Project for University Academic Program Development.
Journal
SmartBot
Method of Research
Experimental study
Subject of Research
Not applicable
Article Title
Highly Dynamic Soft Electrohydraulic Origami Actuators for Agile and Multimodal Robotic Locomotion
Article Publication Date
6-Jan-2026
No comments:
Post a Comment