Self-Sensing Soft Finger Produced from Conductive TPU for Adaptive Grasping and Force Sensing Applications

Soft robotic systems are increasingly being developed for applications that require handling fragile or irregularly shaped objects, such as in agriculture, medical robotics, and industrial automation. This paper presents a self-sensing soft finger fabricated entirely from conductive thermoplastic polyurethane (TPU), combining adaptive grasping capabilities with integrated piezoresistive force sensing. Inspired by the Fin Ray Effect, the finger eliminates the need for external sensors by using its structure for force measurement. Additive manufacturing through Fused Filament Fabrication (FFF) was used to produce the finger, offering an efficient and cost-effective fabrication process suitable for scalable applications. The finger's sensing performance was validated through experimental tests that employed a calibrated load cell to measure applied forces while simultaneously recording the analog output of the self-sensing finger. A Moving Median filter was applied to reduce noise in the analog data, resulting in smoother and more interpretable readings. The results confirmed the reliability of the soft finger under different gripper closure percentages and speeds, demonstrating its capability to track force variations accurately. This study also identifies some challenges, such as inconsistencies in the readings at lower force levels, and outlines possible future improvements, including the integration of multi-material printing to isolate structural elements and enhance sensing differentiation. The results highlight the potential of the self-sensing soft finger for practical applications in delicate object manipulation, particularly in fields like fruit handling, and automated manufacturing systems.