Robotic Arm for Unfastening Screws in Automated Disassembly Process

Abstract

The importance of the Disassembly Process has gradually grown in recent days. The system proposed in this paper is a solution for unfastening screws to help automate disassembly of electronic devices or e-waste using a Robotic Arm. We see a new variety of Electronic Devices coming into market every day, and all these products have short life. They also have uncertainties when it comes to their disassembly after the end of product life. The intention of product disassembly is to minimize value loss at the end of product life. This project makes use of a SCARA robotic arm with 5 DOF to be used for the automatic disassembly of old electronic devices. The idea is to detect the type of the screw from its overhead image as seen by the arm and pick an appropriate driver and unfasten the screw using the driver. This is achieved using the Robot Operating System (ROS), a framework that acts as an meta-operating system to interface the robot. A robotic arm of Scara type was designed for this purpose with a suitable gripper using Autodesk Fusion 360. The arm is capable of detecting and classifying the screws using Yolo V5, a machine learning model.