This paper presents a shear-based management scheme for grasping and manipulating delicate objects with a Pisa/IIT anthropomorphic SoftHand equipped with gentle biomimetic tactile sensors on all 5 fingertips. These ‘microTac’ tactile sensors are miniature variations of the TacTip vision-primarily based tactile sensor, portable cutting shears and may extract exact contact geometry and cordless power shears data at every fingertip to be used as suggestions into a controller to modulate the grasp while a held object is manipulated. Using a parallel processing pipeline, we asynchronously capture tactile photos and predict contact pose and gardening shears force from multiple tactile sensors. Consistent pose and force models across all sensors are developed utilizing supervised deep learning with transfer learning strategies. We then develop a grasp management framework that makes use of contact force suggestions from all fingertip sensors simultaneously, permitting the hand to safely handle delicate objects even below external disturbances. This control framework is utilized to a number of grasp-manipulation experiments: first, retaining a versatile cup in a grasp without crushing it under changes in object weight

Edit

Pub: 16 Aug 2025 18:05 UTC

Views: 14