Design a low-cost delta robot arm for pick and place applications based on computer vision

Abstract

In this paper, we develop a low-cost delta robot arm for grasping objects of unspecified size thanks to a vision system. Stepper motors are used instead of ac servo motors to build a low-cost delta robot arm. Furthermore, we use available materials and machining methods such as laser cutting and 3d printing instead of CNC milling and turning to reduce fabrication costs. The controller is based on a low-cost embedded controller - Arduino Uno for controlling the robot's motion. The vision system is constructed to determine the 3D coordinate of objects in the workspace as well as the sizes of objects. The gripper is opened with a distance of two fingers equal to the size of the objects, and the robot is controlled to the objects' coordinates to grasp them. An application to pick up objects on a conveyor belt is developed to validate the design. The experimental results show that the robot system works correctly, the robot arm moves smoothly, and the information determined by the vision system has a small error, ensuring that the robot can accurately pick up products.