A ROS 2 Based Robotic System to Pick-and-Place Granular Food Materials

Abstract

Granular food material handling remains an unre-solved topic due to the difficulties in determining appropriate grasp configuration. Their quantity is dependent on volume, and grasping a specific quantity is influenced by the food surface in the container, grasp location, and so on. This paper presents a ROS 2 based system for a robotic setup in real food handling applications. An image processing algorithm is proposed to automatically determine the desired grasp position for granular food materials from their container by dividing the surface into several regions. The system can be useful in real applications in food companies, to pick-and-place various food products, pack bento, sort a series of food products from conveyor belt and so on. The proposed system is tested with chopped green onions on a robot setup with a Denso SCARA robotic manipulator and a pneumatic-actuated soft gripper. Several experiments are conducted to demonstrate the effectiveness of the robotic system during pick-and-place tasks, while discussing some limitations that can be addressed in future works.