Page 74 - Read Online

Zhong et al. Soft Sci. 2025, 5, 3 https://dx.doi.org/10.20517/ss.2024.52 Page 13 of 15

on surfaces with joints of different sizes, other sensors of the same type must be calibrated to perform the
same task. Compared to existing soft bending sensors, the DCLS sensor can be used directly without
calibration to accurately sense the bending angle. To validate the directional sensing calibration-free
capability of the sensor in the hand orthotic exoskeleton robot, we conducted an experiment involving the
actuation of the wrist joint. After a volunteer wore the hand orthotic exoskeleton robot, the wrist actuator
assisted in both flexion and extension motions of the wrist joint. Without requiring calibration, the DCLS
sensor demonstrated the ability to discern bending direction variations, as depicted in Figure 6C.
Additionally, the signal stability of the sensor during routine use is also particularly important. We have
determined the variance for each sensor’s signal under these two scenarios (handshaking and knocking) by
analyzing the temporal data provided by the sensors [Supplementary Table 3 and Supplementary Movie 6].
Notably, the findings indicate that, for both cases, the variance measures for signals of all sensors are below
0.5. This outcome substantiates the robustness of the DCLS sensor within hand orthotic exoskeleton
robotics.

CONCLUSIONS
In summary, a novel DCLS bending sensor based on an optical waveguide is proposed and investigated in
this paper. The DCLS sensor can directly sense the bending angles and directions of robot joints of different
sizes without the need of recalibration. It exhibits good measurement accuracy, resolution, and repeatability.
Its sensing capabilities underscore its potential for integration into soft robotics. To showcase its
calibration-free characteristic, we engineered three distinct soft robots catering to diverse applications. For
the soft sorting robot embedded with the DCLS sensor, the sensor can directly measure the bending
curvature of the pneu-nets actuator without calibration, achieving automatic fruit sorting. Furthermore, we
have also successfully integrated this sensor into a fish-inspired robot and a hand orthotic exoskeleton
robot. The results collectively demonstrate the calibration-free characteristic of the sensor’s capability to
measure bending angles and directions. We anticipate numerous promising applications for the DCLS
sensor in various fields, including but not limited to human-machine interactions, clawing robotics, and
human motion monitoring.

DECLARATIONS
Authors’ contributions
Writing - original draft: Zhong, L., Peng, Y.
Methodology: Zhong, L., Tian, X., Wang, J. Y., Wang, J. X., Nei, Z., Chen, X.
Supervision and funding acquisition: Peng, Y.
Writing - review and editing: Peng, Y.
All authors reviewed and commented on the manuscript before publication.

Availability of data and materials
The data presented in this study are available upon request from the corresponding author.

Financial support and sponsorship
This research was supported by the National Natural Science Foundation of China (No. 52105563, No.
52475573), the Natural Science Foundation of Zhejiang Province (No. LTGY23E050002), the National Key
Research and Development Program of China (No. 2023YFC2811500), Zhijiang Youth Project of Zhejiang
Provincial Philosophy and Social Science Planning (No. 22ZJQN14YB), the Key Research and Development
Programme of Zhejiang (No. 2023C03196, No. 2022C03103, and No. 2024C03259), and the Fundamental
Research Funds for the Central Universities.