Analysis of Intelligent Force Control Methods for Red-Meat Gripping Applications

Abstract

Slaughterhouse automation is a challenging but important effort that could potentially improve the efficiency and hygiene of meat-processing while replacing human operators with machines and robots in dangerous and unpleasant jobs. Automated handling of carcasses require high level of intelligence, autonomy and adaptability, thus AI-driven systems, various sensors and intelligent tools are necessary parts of such systems. This paper presents novel force control methods implemented in a smart gripper intended for automated pig cutting. While the gripper is capable of handling high payloads (whole legs and 15–20 kg of intestines), the force control algorithm also allows delicate grasping on soft tissues. The gripping force is estimated separately for the two fingers of the gripper based on novel timing belt strain measurement methods. The elongation of the belts are measured by strain-gauges, and by magnetic encoders measuring the rotations of axes. The force sensors were tested with a tensile testing machine, and the functioning force-controlled gripper was tested with real pig carcasses too. The sensors have also proved useful for giving other feedback during automated pig cutting too, such as warning about inadequate gripper positioning relative to the target, and error signal generation when the target object slipped out of the gripper.