AGV indoor localization: a high fidelity positioning and map building solution based on drawstring displacement sensors
Shih-Yuan Wang,Che-Ming Li,Sze-Teng Liong,Yu-Ting Sheng,Yen-Chang Huang,Gen-Bing Liong,Yee-Siang Gan
DOI: https://doi.org/10.1007/s12652-024-04755-5
IF: 3.662
2024-02-20
Journal of Ambient Intelligence and Humanized Computing
Abstract:The rapid evolution of artificial intelligence (AI) and robotics technologies are bringing drastic changes to society and industry in these recent years. The impressive progress in facilitating smart manufacturing in this era of Industry 4.0 has made our lives much more convenient than ever before. Despite the greater reliability and stability of the robotic system, there are several challenges to overcome such as the restrictions of scenes, obstacles, and hardware specifications. Since a high precision positioning algorithm is of paramount importance in devising a mobile robot, the research in developing simultaneous localization and mapping has been garnered immerse attention especially from domains of the computer vision and autonomous robots. In this paper, a novel method is presented to significantly enhance the positioning precision of indoor unmanned guided vehicles. The approach involves several steps, including setting up hardware configurations and collecting relevant data by installing necessary devices and system packages within the robot operating system (ROS). Trilateration is employed to determine the relative position of the mobile robot using distance measurements. Coordinate transformation is then conducted to convert the collected input data of relative distances and orientations. Trajectory paths are obtained, and occupancy maps are constructed to estimate the resulting trajectory and generate a 2D grid map. Indoor localization and mapping are achieved using three drawstring displacement sensors along with orientation information from an Inertial Measurement Unit (IMU). The proposed method is extensively evaluated through experimentation on predefined navigation paths, and its performance is compared to state-of-the-art methods such as RealSense T265, Hector SLAM, and wheel odometry. The results show that the proposed method exhibits compelling performance in both mean error and occupancy map construction. Ultimately, the findings reported herein offer interesting insights and shed light on an alternative solution in introducing a robust positioning system.
computer science, information systems,telecommunications, artificial intelligence