Affordable 3D-printed miniature robotic gripper with integrated camera for vision-based force and torque sensing

Gripping forces of compliant robotic grippers can be estimated from finger deformations (e.g., by a camera) and used to prevent object damage (e.g., by closed-loop force control). We present a proof-of-concept compliant gripper with two novelties: (1) a disposable finger part, 3D-printed in a single step and integrating a miniature camera; (2) a method requiring minimal training data to estimate gripping force and six-axis force/torque. Force/torque estimates agreed well with a reference sensor in a controlled lab setting. Application studies demonstrated our gripper’s potential to handle rigid and soft objects while applying visual servoing and force thresholding. Although variable lighting and nonuniform objects required more careful and repeated calibration and alignment, we consider higher robustness achievable with more expensive deformation tracking or force models. While our gripper remains at an early proof-of-concept stage, after these improvements it could be applicable to several manipulation tasks benefiting from force sensing and versatility.