Xiaosa Li

Xiaosa Li

Auto-generated profile, based on publicly available information

Xiaosa Li

Tsinghua Shenzhen International Graduate School, Tsinghua University, Shenzhen 518055, P. R. China

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
12 September 2025 in Science Advances
Verified
Journal Article
Open access
XLXiaosa Li
Xiaosa Li
Xiao XiaoXiao Xiao
Xiao Xiao
ZLZixiao Liu
Zixiao Liu
JGJunhao Gong
Junhao Gong
ZLZenan Lin
Zenan Lin
CWChangsheng Wu
Changsheng Wu
Embodied intelligence in soft robotics offers unprecedented capabilities for operating in uncertain, confined, and fragile environments that challenge conventional technologies. However, achieving true embodied intelligence—which requires continuous environmental sensing, real-time control, and autonomous decision-making—faces challenges in energy management and system integration. We developed deformation-resilient flexible batteries with enhanced performance under magnetic fields inherently present in magnetically actuated soft robots, with capacity retention after 200 cycles improved from 31.3 to 57.3%. These compliant batteries enable large-area deployment of 44.9% across the robot body, and their vertical integration with rationally designed flexible hybrid circuits minimizes additional stiffness while maintaining deformability. This actuator-battery-sensor vertical integration methodology maximizes functional area utilization in a manta ray–inspired soft robot, establishing an untethered platform with sensing, communication, and stable power supply. The system demonstrates embodied intelligence in aquatic environments through diverse capabilities including perturbation correction, obstacle avoidance, and temperature monitoring, with proprioceptive and environmental sensing enabling real-time decision-making during magnetically actuated locomotion.

14 July 2022 in IEEE Robotics and Automation Letters
Verified
Journal Article
Closed access
XLXiaosa Li
Xiaosa Li
CLChengyue Lu
Chengyue Lu
ZSZiwu Song
Ziwu Song
Wenbo DingWenbo Ding
Wenbo Ding
XZXiao-Ping Zhang
Xiao-Ping Zhang
Magnetic actuation system manipulates micro soft or rigid robots by a controllable magnetic field to move them freely in the narrow or enclosed space, which has demonstrated its huge potential in medical interventional surgery and drug delivery. However, the limited working space of paired or area-centered electromagnets restricts its practical applications. In this paper, we propose a convenient coils drive scheme for the scalable electromagnet array, and present an efficient planar magnetic actuation system with a spacious workspace. During the actuation process, our system activates selectively the effective electromagnets neighboring to the magnetic robot by coil selectors, and generates an alternating magnetic field with sufficient gradients to guide the robot's orientation and position. For the soft magnetic pipe, our system can push it to perform the continuous deflections around the stand columns on the plane. For the rigid magnetic cube, the designed magnetic-quadrupole structure allows it to receive various forces from different directions, and achieve a stable displacement in the heterogeneous magnetic field.

27 June 2024 in IEEE Sensors Journal
Verified
Journal Article
Closed access
ZSZiwu Song
Ziwu Song
CLChenchang Li
Chenchang Li
ZQZhentan Quan
Zhentan Quan
SMShilong Mu
Shilong Mu
XLXiaosa Li
Xiaosa Li
XZXiao-Ping Zhang
Xiao-Ping Zhang
Adapting to various terrains and improving locomotion performance have always been challenging for legged robots. Inspired by the animal somatosensory system in walking and balancing, the robot can benefit from the foot sensors as well when interacting with the terrain and gain better performance. In this paper, we propose a visuo-tactile foot sensor with embedded markers, named TacTID, for legged robot terrain identification. By introducing elastic embedded markers on an elastic hemispherical shell, TacTID could detect 3-dimensional pressure distribution directly, based on which, the terrain coefficient of sliding friction (COSF) and terrain effective stiffness can be estimated precisely. The results of the COSF prediction experiment demonstrate that TacTID exhibits an average error of 0.036, which is notably lower than that of TacTip. For the effective stiffness estimation experiment, TacTID is tested on 10 silicone cubes with different Young’s moduli, and the absolute average estimation error is less than 0.1 MPa. Besides, a quadruped robot is equipped with TacTID for real-time terrain identification, and the possibility to estimate COSF and the terrain effective stiffness in various terrains is demonstrated. This research highlights the potential of TacTID for robotic applications requiring precise contact force and terrain identification in real time.

16 October 2024 in IEEE Robotics and Automation Letters
Verified
Journal Article
Closed access
XLXiaosa Li
Xiaosa Li
RZRunze Zhao
Runze Zhao
XCXinyue Chai
Xinyue Chai
ZWZimo Wang
Zimo Wang
QTQianqian Tong
Qianqian Tong
Wenbo DingWenbo Ding
Wenbo Ding
Cutaneous haptics is helpful to tame the human-machine mismatch by interactive tactile feedback and perform precise manipulations for virtual immersive interactions. However, wearable tactile gloves cover the palm and fingers greatly, thus reducing the tactile information from interactive objects and causing the inconsistency between virtual and practical operations. In this work, we design a lightweight finger-worn vibration bot, named VibroBot, to provide the individual vibrotactile feedback to each finger without compromising the hand dexterity. Each VibroBot, weighing only 2.9 grams, integrates components of the power, a wireless chip and a coil actuator, to receive programmable waveform signals and perform the real-time vibration feedback wirelessly on finger. Our design features six rapidly distinguishable vibration modes with the 96.4% recognition rate in 2.0s, each guiding one of three finger joints for flexing or extending to a specified angular range. When worn on all five fingers, VibroBots can collaboratively guide the user with multi-dimension semantics, to adjust five fingers for target hand gestures at the same time. In virtual gesture training experiments, VibroBots were used to correct users' muscle memory bias for common gestures, and reduced the average gesture error from about 30 $^\circ$ to less than 15 $^\circ$ . Haptic guidance by VibroBots shows its potential in various Tactile Internet scenarios that require a large number of pretrained hand operations, such as robotic teleoperation and virtual surgical training.