Vision-based manipulation of transparent plastic bags in industrial setups.

Abstract

Introduction: This paper addresses the challenges of vision-based manipulation for autonomous cutting and unpacking of transparent plastic bags in industrial setups, contributing to the Industry 4.0 paradigm. Industry 4.0, emphasizing data-driven processes, connectivity, and robotics, enhances accessibility and sustainability across the value chain. Integrating autonomous systems, including collaborative robots (cobots), into industrial workflows is crucial for improving efficiency and safety.

Methods: The proposed system employs advanced Machine Learning algorithms, particularly Convolutional Neural Networks (CNNs), for identifying transparent plastic bags under diverse lighting and background conditions. Tracking algorithms and depth-sensing technologies are integrated to enable 3D spatial awareness during pick-and-place operations. The system incorporates vacuum gripping technology with compliance control for optimal grasping and manipulation points, using a Franka Emika robot arm.

Results: The system successfully demonstrates its capability to automate the unpacking and cutting of transparent plastic bags for an 8-stack bulk-loader. Rigorous lab testing showed high accuracy in bag detection and manipulation under varying environmental conditions, as well as reliable performance in handling and processing tasks. The approach effectively addressed challenges related to transparency, plastic bag manipulation and industrial automation.

Discussion: The results indicate that the proposed solution is highly effective for industrial applications requiring precision and adaptability, aligning with the principles of Industry 4.0. By combining advanced vision algorithms, depth sensing, and compliance control, the system offers a robust method for automating challenging tasks. The integration of cobots into such workflows demonstrates significant potential for enhancing efficiency, safety, and sustainability in industrial settings.