Soft suction cup senses forces to handle kiwifruit gently
Mechanical Interaction Sensing for Robotic Kiwifruit Handling with an Optoelectronic Soft Suction Cup
Robotics
Summary
Handling delicate fruits like kiwifruit without damage is a challenge for robots. The authors designed a soft suction cup with built-in light sensors to measure how much force it applies when gripping fruit. Their system can estimate forces very accurately, even on fruits it hasn’t seen before, helping robots know when they are squeezing too hard or too softly. This could help robots handle fruit more gently and reduce waste.
What this means in practice
- •For robotic fruit pickers: Measure forces during fruit gripping to reduce damage when harvesting kiwifruit with robots.
- •For soft robotics engineers: Incorporate optical sensing in soft suction cups to reliably estimate contact forces for delicate object handling.
Authors
Romina Adhami, Ebrahim Shahabi, Alessio Mondini, Samuel M. Youssef, Cosimo Della Santina, Gualtiero Volpe, Barbara Mazzolai
Abstract
Robotic handling of delicate fruit benefits from compliant contact and quantitative sensing of the mechanical interaction at the end-effector. This study investigates an optoelectronic soft suction cup with a kiwifruit-specific bicomponent design for estimating interaction forces during contact with real fruit. During known-contact intervals, the optical signals accurately estimate interaction force on fruits excluded from model training: the best shallow network achieves grouped out-of-fold pooled R-squared = 0.968 +/- 0.005 (MAE = 0.138 N), with a median per-fruit held-out R-squared = 0.973 (MAE = 0.134 N). A simple monotonic mapping from the mean optical response alone achieves R-squared = 0.925 +/- 0.005, showing that substantial force-related information is already present in the optical response. Leave-one-collection-day-out validation remains strong (R-squared = 0.955 +/- 0.031), supporting robustness across acquisition sessions. As an exploratory secondary analysis, binary Soft/Firm classification from compression-test reference labels achieves moderate balanced accuracy (approximately 0.705) under grouped held-out-fruit validation, while finer three-class grading shows limited performance.