2019/09/16 by Artūras Straižys, Straižys, Artūras, Michael Burke +3 · 1 citation
Engineering · #Soft Robotics and Applications #Robot Manipulation and Learning #Robotic Mechanisms and Dynamics
paper · pdf · doi:10.48550/arxiv.1909.07247
Precision cutting of soft-tissue remains a challenging problem in robotics,\ndue to the complex and unpredictable mechanical behaviour of tissue under\nmanipulation. Here, we consider the challenge of cutting along the boundary\nbetween two soft mediums, a problem that is made extremely difficult due to\nvisibility constraints, which means that the precise location of the cutting\ntrajectory is typically unknown. This paper introduces a novel strategy to\naddress this task, using a binary medium classifier trained using joint torque\nmeasurements, and a closed loop control law that relies on an error signal\ncompactly encoded in the decision boundary of the classifier. We illustrate\nthis on a grapefruit cutting task, successfully modulating a nominal trajectory\nfit using dynamic movement primitives to follow the boundary between grapefruit\npulp and peel using torque based medium classification. Results show that this\ncontrol strategy is successful in 72 % of attempts in contrast to control using\na nominal trajectory, which only succeeds in 50 % of attempts.\n