← Back ICRA 2026

Acoustic Sensing for Universal Jamming Grippers

Lion-Constantin Weber, Theodor Marius Wienert, Martin SplettstÃ¶Ãer, Alexander Koenig, Oliver Brock

PDF

AI summary

Key figure (auto-extracted from paper)

Acoustic sensing embedded in a universal jamming gripper preserves compliance while accurately perceiving object size, orientation, material, and class without compromising grasping performance.

acoustic sensing universal jamming grippers morphological sensing soft robotics tactile perception machine learning

Problem

Traditional tactile sensors compromise the compliance of universal jamming grippers, reducing their grasping performance, yet tactile sensing is essential for unstructured environments.

Approach

The gripper's soft body acts as a sensor by transmitting sound from an internal speaker to a microphone; machine learning decodes the modulated acoustic signals to reconstruct object properties.

Key results

2.6 mm error in object size estimation
0.6° error in orientation prediction
Up to 100% accuracy in material discrimination
85.6% accuracy in classifying 16 everyday objects

Why it matters

Enables robust, high-resolution tactile perception for soft grippers in unstructured environments without sacrificing their inherent compliance or grasping ability.

Abstract

Universal jamming grippers excel at grasping unknown objects due to their compliant bodies. Traditional tactile sensors can compromise this compliance, reducing grasp- ing performance. We present acoustic sensing as a form of morphological sensing, where the gripper’s soft body itself becomes the sensor. A speaker and microphone are placed inside the gripper cavity, away from the deformable membrane, fully preserving compliance. Sound propagates through the gripper and object, encoding object properties, which are then reconstructed via machine learning. Our sensor achieves high spatial resolution in sensing object size (2.6 mm error) and orientation (0.6◦error), remains robust to external noise levels of 80 dBA, and discriminates object materials (up to 100 % accuracy) and 16 everyday objects (85.6 % accuracy). We validate the sensor in a realistic tactile object sorting task, achieving 53 minutes of uninterrupted grasping and sensing, confirming the preserved grasping performance. Finally, we demonstrate that disentangled acoustic representations can be learned, improving robustness to irrelevant acoustic variations.

Index terms

Soft Sensors and Actuators Perception for Grasping and Manipulation Force and Tactile Sensing