Indoor Positioning of Autonomous Guided Vehicles Using LiDAR-Corner Cubes with Adaptive Monte Carlo Localization

Abstract

In the context of Industry 4.0, integrating automation equipment into production lines has become increasingly prevalent. The efficiency of a factory’s production is significantly influenced by the autonomous handling and supply of materials or workpieces. This study focuses on the Autonomous Guided Vehicle (AGV) system, which utilizes advanced technologies such as laser optical radar (Light Detection and Ranging, LiDAR) and reflection ridges (Corner Cubes) to achieve indoor positioning and synchronous construction of the contour in a spatial field guided by a specifically dedicated map. The primary emphasis is on AGV indoor positioning technology, employing LiDAR and reflections to calibrate each position coordinate. The research establishes the core technology of indoor Simultaneous Localization and Mapping (SLAM) through the application of a Robots Operating System (ROS), which performs simulation, testing, and the verification analysis of the AGV mechanism. The study also develops a dynamic model for the AGV system, estimates optical position parameters, and integrates them into the Adaptive Monte Carlo Localization (AMCL) combined with the optical indoor positioning algorithm. The hybrid AMCL-Optical positioning provides superior accuracy than individual methods.