A Soft Grasper with Bioinspired Morphology and Synthetic Nervous System Control Reduces Damage to Deformable Objects and Fruits During Handling

The design of robotic graspers that can safely interact with deformable, damage-prone materials such as fruits, vegetables, and biological tissues remains an ongoing challenge in robotics. Conventional robotic graspers made of mostly rigid materials have limited compliance and tactile sensing, reducing their applicability to contact-rich manipulation of soft objects. In contrast, humans and animals can interact with their environments safely and intelligently through their bodies' structural properties and nervous systems' computational capabilities. In this article, we present the design and control of a soft grasper inspired by the sea slug, Aplysia californica, and compare its performance with rigid graspers. The soft jaws and actuators allow the grasper to mimic Aplysia's force sensing capability and its ability to conform to complex food as it grasps. Combining Synthetic Nervous Systems (SNSs), an artificial neural network model inspired by computational neuroscience, and network architectures inspired by Aplysia's feeding control circuitry, we designed distributed and interpretable pick-and-place controllers for the soft grasper and its rigid counterparts. During grasping, these controllers either command a fixed closure radius (feedforward position control) or cap the contact force at a predefined level (force feedback control). We first validated our approach in simulation, demonstrating that the controllers can perform pick-and-place behavior that is robust to sensor noise. We then extended the validation to the physical platform to quantitatively compare how much deformation these graspers induced on soft objects. Fruits such as strawberries, tomatoes, and avocados showed little deformation after they were handled by the soft grasper, suggesting that this approach might have significant agricultural uses. The experimental data suggest the value of the bioinspired soft grasper for soft object manipulation.