Hybrid Vision-Force Robot Force Control for Tasks on Soft Tissues

Abstract

For automation tasks involving robotic manipulator that contacts with soft tissue such as massage and robotic surgery, there are challenges posed by non-linearity, dynamic stiffness of soft tissue, and differences in the mechanical properties of the individual. This study presents a hybrid vision-force control approach for tasks on soft tissue. A vision sensor is used to measure soft tissue shape and calculate desired orientation when the end-effector interacts with soft tissue. The orientation controller based on feed-forward plus feedback control is designed to drive the end-effector track desired orientation. Adjustable variable impedance control is proposed to deal with non-linear and dynamic stiffness of soft tissue; by adjusting the parameter of the controller online, an ideal normal force tracking can be reached. Experiments show that the combination of the above two controllers can achieve great force control on soft tissue.