• Grasp-based product mix enables modular soft gripper design for food handling • Human grasp data used to optimize soft robotic fingers via kinematic modeling • Cable routing and joint stiffness tuned for trajectory accuracy and low force • 3D-printed soft fingers achieve 97% match to recorded human motion trajectories • Modular gripper swaps in 5 minutes, ideal for high-mix food automation systems Robotic automation in the food industry remains challenging because of the wide variability in product shape, texture, and fragility, which demands adaptable and efficient grasping strategies. This study presents an integrated methodology for the design and optimization of soft robotic grippers, centred on three complementary frameworks that collectively enhance adaptability, performance, and design efficiency. The first framework introduces a grasp-based product mix concept that classifies food items according to their required grasp type, enabling a systematic mapping between product characteristics and grasp strategies. The second framework establishes a streamlined design pipeline that combines human demonstration data, grasp strategy optimization, and kinematic modelling to identify effective grasp configurations while reducing development time and cost. The third framework defines an optimization process that incorporates hyperelastic material behaviour, optimized cable routing paths, and joint stiffness tuning to reduce actuation effort and improve grasp stability. The methodology was validated through the development of modular gripper prototypes implementing three representative grasp strategies: power, pinch, and hook. Across 15 diverse food items, the optimized fingers achieved trajectory accuracies exceeding 97%, while mix-specific grasp success rates reached 100% for power and hook configurations and 83% for pinch. The digital redesign phase required approximately 5 minutes per configuration, full gripper fabrication was completed within 24 hours, and configuration switching was achieved in under 5 minutes. Rather than pursuing a universal gripper, this work demonstrates a scalable, task-driven framework for rapidly developing specialized soft robotic grippers, providing a practical pathway toward adaptable automation in food processing, packaging, and high-mix manufacturing environments.
Matte et al. (2026) studied this question.