Development Of Vision-Guided Pallet Engagement System

Abstract

In recent years, the rapid advancement of automation technologies has significantly transformed industrial material handling systems, reducing human dependency while improving efficiency and safety. This paper presents a compact autonomous pallet docking system designed for intelligent warehouse applications. The proposed system utilizes a pre-mapped environment generated using LiDAR-based SLAM and operates using the ROS 2 framework. The robot autonomously navigates from a predefined home position, detects pallets using April Tag-based vision, and performs precise docking through centroid estimation. A scissor lift mechanism integrated with limit switch feedback ensures controlled pallet lifting, while an ultrasonic sensor enhances obstacle avoidance during load transportation. The system dynamically switches sensing strategies to ensure safe operation during both docking and undocking operations. The complete workflow from navigation to pallet docking is implemented and validated in a real-time environment. The proposed system demonstrates a cost-effective and scalable approach for autonomous pallet handling in smart industrial environments.