Summary: One of the hardest things for a robot to do isn’t lifting a heavy box—it’s clicking a computer mouse or playing a piano key. While robots have mastered “the grip,” they historically lack proprioception, the internal sense of where their fingers are in space.
A collaborative research team has solved this by developing a “rigid-soft” hybrid hand equipped with omnidirectional bending sensors. The study demonstrates a robotic hand that can “feel” its own pitch and yaw, allowing it to perform delicate tasks like using scissors with human-like precision.
Key Facts
- 18 Degrees of Freedom: The hand features 18 active joints, closely mimicking the complexity of a human hand’s range of motion.
- Optical Sensing: Each finger uses segmented PMMA fibers and trichromatic LEDs (Red, Green, Blue). By measuring how different colors of light fade as the fiber bends, the robot can distinguish between folding a finger (pitch) and moving it side-to-side (yaw).
- Decoupling Motion: Unlike previous soft sensors that get “confused” when a finger moves in two directions at once, this system separates the signals, maintaining a measurement error of only ±2.13°.
- Proven Stability: The hand successfully demonstrated closed-loop control in three high-coordination tasks: playing the piano, operating a mouse, and cutting with scissors.
Source: AIRCAS
A new soft sensing system could help humanoid robots move their hands with far greater precision in delicate, human-like tasks. The study introduces a dexterous robotic hand equipped with omnidirectional bending sensors that can track both pitch and yaw at the finger joints, allowing the system to perceive complex finger posture in real time.
By combining flexible sensing with a rigid-soft hand design, the researchers created a platform that not only moves more naturally but also performs demanding actions such as using scissors, operating a mouse, and playing the piano with improved control and stability.
Robotic hands have made major progress in grasping and pinching, but many still struggle with the finer motions that make the human hand so versatile. One key limitation is proprioception: while human fingers constantly sense their own position and movement, most humanoid hands remain weak at perceiving posture across multiple degrees of freedom.
Existing soft sensors often detect only one bending mode or suffer from coupling problems when fingers flex and move sideways at the same time. This leaves a gap between robotic grasping and true dexterous manipulation. Based on these challenges, deeper research was needed into soft sensing systems capable of decoupling and accurately tracking multidirectional finger motion.
Researchers from Zhejiang University, Hangzhou Dianzi University, and Lishui University reported the work in Microsystems & Nanoengineering in 2026. The study presents a humanoid dexterous hand designed to solve a central problem in advanced robotics: how to give robot fingers a reliable sense of their own posture during complex motion.
By embedding a new omnidirectional soft bending sensor into the hand, the team enabled real-time perception of both flexion and side-to-side movement in delicate manipulation tasks.
The hand features 18 active degrees of freedom and five rigid-flexible fingers, with each finger integrating a soft optical sensor built from segmented PMMA fibers, a trichromatic LED, and a chromatic detector.
The design works by tracking how red, green, and blue light attenuate differently as the sensor bends. Because the fiber layout separates responses to pitch and yaw, the system can decouple the two motions instead of mixing them together.
The paper reports strong repeatability over 100 cycles, with RMSE values of 2.1%, 1.9%, and 3.2% across the three optical channels. Under single bending, the average measurement error was only ±2.13° for pitch and ±2.34° for yaw. Crosstalk remained low: pure yaw contributed 3.2% to pitch, while pure pitch contributed 4.1% to yaw, with signal-to-crosstalk ratios of 50.68 dB and 30.81 dB, respectively.
The team then moved beyond bench testing and demonstrated the hand in three visually compelling tasks—cutting with scissors, clicking a mouse, and playing piano keys—showing closed-loop posture control in actions that require subtle coordination rather than simple gripping.
The researchers suggest that the real advance is not just a new sensor, but a new way of giving robotic hands a more human-like internal awareness of motion. In their conclusion, they emphasize that the integrated rigid-soft design supports natural movement, while the sensing system delivers the stability, repeatability, and multi-DoF posture perception needed for complex operations. That combination could make future humanoid hands more capable in tasks where precision matters most.
This work points toward robotic hands that are not only stronger or faster, but more skillful. Better posture perception could improve humanoid robots used in service settings, industrial assembly, rehabilitation devices, and other environments where fingers must adapt to fragile or highly varied objects.
The study’s demonstrations also hint at broader possibilities in human-robot interaction, where smoother and safer hand motion is essential. By showing that soft optical sensing can remain accurate while enabling complex, multidirectional motion, the research moves robotic manipulation closer to the responsiveness and finesse of the human hand.
Funding Information
This research was supported by the National Natural Science Foundation of China (No. 52475573), the Natural Science Foundation of Zhejiang Province (No. LTGY23E050002), the National Key Research and Development Program of China (No. 2023YFC2811500), the Science and Technology Innovation Project of the General Administration of Sport of China (24KJCX074), the Key Research and Development Programme of Zhejiang (No. 2024C03259, No. 2023C03196), and the Fundamental Research Funds for the Central Universities.
Key Questions Answered:
A: Cutting requires a constant, subtle adjustment of “side-to-side” (yaw) and “up-and-down” (pitch) pressure. Most robots only sense one direction, so the scissors slip or jam. This new hand uses “omni-sensors” that track both movements simultaneously, allowing the robot to adjust its grip just like a human would.
A: Think of it like a glowing straw. When you bend the straw, the light inside dims. By using three different colors of light (RGB) and specific fiber layouts, the sensors can tell exactly how the straw is bending based on which color dims the most. This “optical crosstalk” is then decoded into a precise finger position.
A: Absolutely. Because the sensors are “soft” and integrated into a flexible design, they are much more durable and natural-feeling than traditional rigid sensors. This could lead to prosthetic limbs that allow users to perform fine-motor tasks, like typing or playing an instrument, with much higher accuracy.
Editorial Notes:
- This article was edited by a Neuroscience News editor.
- Journal paper reviewed in full.
- Additional context added by our staff.
About this robotics and neurotech research news
Author: Yuan Wang
Source: AIRCAS
Contact: Yuan Wang – AIRCAS
Image: The image is credited to Neuroscience News
Original Research: Open access.
“Soft sensor for omnidirectional posture perception in humanoid dexterous hands” by Liang Zhong, Xiaoqing Tian, Jiyong Wang, Xian Song, Jianfeng Li & Yuxin Peng. Microsystems & Nanoengineering
DOI:10.1038/s41378-026-01179-3
Abstract
Soft sensor for omnidirectional posture perception in humanoid dexterous hands
This study presents the development of a novel omnidirectional soft bending sensor tailored for humanoid dexterous hands to facilitate posture perception in delicate manipulation tasks.
Drawing inspiration from the human hand’s intricate design and proprioceptive capabilities, this study aims to enhance the dexterity of robotic hands, particularly in multi-degree-of-freedom (DoF) motion and posture perception.
To this end, we designed a humanoid dexterous hand featuring 18 active DoFs, with five rigid-flexible structured fingers for improved joint mobility. Each finger is equipped with our innovative omnidirectional bending sensor, utilizing segmented polymethylmethacrylate (PMMA) optical fibers, a trichromatic LED, and a chromatic detector to detect the pitch and yaw angles of the metacarpophalangeal joints.
The sensor demonstrated excellent measurement performance, stability, and repeatability in challenging tasks such as using scissors, operating a computer mouse, and playing the piano.
This technology addresses the challenges associated with multi-DoF motion and omnidirectional posture perception in robotic hands, thereby enhancing their capabilities in delicate manipulation tasks and paving the way for further advancements in humanoid dexterous hand development.
