A Proprioceptive Soft Tentacle Gripper Based on Crosswise Stretchable Sensors

• Published in: IEEE/ASME transactions on mechatronics Vol. 25; no. 4; pp. 1841 - 1850
• Main Authors: Xie, Zhexin, Yuan, Feiyang, Liu, Zemin, Sun, Zhaoning, Knubben, Elias M., Wen, Li
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.08.2020; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Actuation; Actuators; Bending; Biomimetics; Detection; Elongation; Grasping (robotics); Modeling and control; Modelling; Object recognition; Robot sensing systems; Robotics; Robots; Sensor phenomena and characterization; Sensors; soft gripper; Soft robotics; soft sensor; Strain; Suction
• ISSN: 1083-4435; 1941-014X
• DOI: 10.1109/TMECH.2020.2993258

Soft robotics has attracted growing attention due to its compliant structure, safe operation, and promising applications in constrained and unstructured environments. Octopus can employ its arm and suckers to catch prey in complex environments. Intrigued by these capabilities, we explored a new approach in sensing, modeling, and control of a bioinspired soft robot that enables both bending and suction for manipulating and perception of environmental load. By implementing EGaIn-based soft sensors with a crosswise manner on the soft tentacle, both sensory feedbacks of the elongation and expansion induced by the pneumatic inflation can be acquired. In this article, we built a variable curvature model for the bending kinematic reconstruction of the soft robot based on the feedback of the soft sensors. Both modeling and experimental results show that this crosswise sensor design allows for more accurate sensing of the soft robot's bending than the one-axis design. With the crosswise sensor, the robotic tentacle can distinguish the bending under external stimuli and internal self-actuation. We experimentally verified the sensory response as well as the accuracy of the self-sensing system and the model under freeload bending, externally loaded bending, and grasping conditions. By synergistically adopting the sensing of the vacuum-actuated suckers with the bending arm, we show that the soft robot can detect, suck, and bend to grasp objects. The results from this article may provide new design and application insights into the creation of autonomous soft manipulators with both bending and suction for tasks in constrained environments.