Industry News Dispatch | AI Robotics (II): Cutting-Edge Technology Updates [2025.8.1-2025.8.31]

Industry News DispatchAI Robotics

Gain insights into the forefront of technology and seize opportunities. This column is continuously updated to provide you with the latest industry dynamics, featuring selected products, industry news, technological advancements, policy updates, industry perspectives, and investment trends, serving as your professional industry intelligence officer!

Current report period: 2025.8.1-2025.8.31

Technological Advancements

(1) UCBerkeleyBAIRLaboratory has released aReinforcement Learning Framework — Human-in-the-Loop Sample-Efficient Robotic Reinforcement Learning (HIL-SERL), which can acquire a wide range of precise and dexterous robotic skills, achieving a 100% success rate across all tasks with only 1-2.5 hours of training, far exceeding the baseline method’s average success rate of less than 50%.

(2) Princeton University‘s research team proposed theReality Promises concept and developed a hybrid reality interaction system based oninvisible robots. Users can select objects through gestures, while the invisible robots perform the actual operations, achieving a seamless transition from virtual commands to physical actions.

(3) Peking University‘s Advanced Manufacturing and Robotics Institute’s Intelligent Bionic Design Laboratory, led by Xie Guangming, proposed a novel manufacturing method based onintegrated programmable fabric stacking, successfully developing a series of multifunctionalbionic soft robots.

(4) University of Science and Technology of China and several other universities jointly developed amagnetorheological (MR) actuator-based hand exoskeleton system (MRHE), which significantly enhances the user’s hand grip endurance and strength, utilizing the body’s own energy for actuation and energy storage without relying on external power sources.

(5) Wuhan University of Technology, Beijing Institute of Technology, Peking University People’s Hospital, and The Chinese University of Hong Kong collaborated to innovatively propose agrating area variable diameter FBG sensing microstructure for achieving high-precision perception ofsurgical robot and tissue interaction forces.

(6) On August 15, Peking University developed a method based on deep reinforcement learning totrain a jellyfish-like swimming robot to efficiently track moving targets in a two-dimensional flow field. This work expands the application of machine learning in the field offlexible object control in fluid environments.

(7) Zhejiang University proposed a“deformation-driven” strategy based on free radical diffusion, breaking through the geometric constraints on LCE actuators. This method is applicable to most acrylate LCE systems and is expected to promote the large-scale customization of actuators andsoft robots.

(8) Suzhou Institute of Nano-Tech and Nano-Bionics developed a new type of elastic artificial muscle fiber, which is apolydimethylsiloxane (PDMS)-wrapped nylon yarn elastic artificial muscle fiber that exhibits elastic driving performance similar to biological muscles, combining actuation and sensing, providing new directions forrobotics and wearable technology.

(9) Xihua University, inspired by the contraction mechanism of insect muscles, developed aelectromagnetic elastomeric drive mechanism (Elasto-Electromagnetic mechanism, EEM) that utilizes the balance of elastic force and static magnetic attraction to achieve muscle-like contractions in robots, allowingmicro-robots to autonomously crawl, swim, jump, and navigate complex environments, bringing new possibilities for search and rescue, exploration, and medical fields.

(10) On August 13, Harbin Engineering University published research onelectro-hydraulic driven deep-sea soft robots in the top international journal Science Robotics, proposing an innovative“flexible electro-hydraulic unit liquid-solid plasticization mechanism” and “integration of electro-hydraulic and plasticizing media” strategy, transforming “electro-induced flow” into aflexible driving mechanism suitable for extreme deep-sea environments.

(11) On August 7, Ocean University of China collaborated withNorth Carolina State University to publish the latest results in the journal Advanced Materials Technologies, developing a novelpiezoelectric driven and sensing integrated soft robot that can move quickly and flexibly, and can identify the surrounding environment through “sensing vibrations” like a spider, achieving a high degree of integration of actuation and sensing.

(12) Harvard University researchers invented a newsoft wearable robotic device that can respond to each user’s precise movements, providing more personalized assistive functions and better, more controllable support for users’ daily tasks. This research was published in Nature Communications.

(13) Beihang University developed China’s firstmicro-drone with autonomous hovering capability that can search for signs of life in disaster debris, monitor wildlife up close without disturbance, and navigate through narrow pipes that are difficult for humans to access for inspections.

(14) State University of New York at Buffalo, inspired by the mechanisms of human tactile neurons, developed aself-driven tribovoltaic effect-based bionic sliding sensing system (BTS) with self-powering capabilities, high-sensitivity sliding perception, and real-time feedback control functions, achieving high-sensitivity sliding detection and feedback adjustment without external power supply, with a system response time as low as milliseconds, far superior to visual-tactile detection principles. This provides a new approach for future intelligent robot grasping, medical assistance, and human-machine interaction interfaces.

END

Note: This industry news content is compiled from public reports.

The content provided by this public account is for personal communication, learning, and research purposes only and may not be used for other commercial purposes. If there are any content, copyright, or other issues, please leave a message. Please indicate the author and source when reprinting.

Guangdian Yuntong Research Institute

Insight into AI, foresee the future!

👇 Share, Save, View, and Like 👍