Multi-sensor fusion for automated guided vehicle positioning

• Main Author: Ahmed Elshayeb, Said Abdo
• Format: Thesis
• Language: English; English
• Published: 2010
• Subjects: TJ Mechanical engineering and machinery; TJ210.2-211.47 Mechanical devices and figures. Automata. Ingenious mechanisms. Robots (General)
• Online Access: http://eprints.uthm.edu.my/3600/1/24p%20SAID%20ABDO%20AHMED%20ELSHAYEB.pdf; http://eprints.uthm.edu.my/3600/2/SAID%20ABDO%20AHMED%20ELSHAYEB%20WATERMARK.pdf; http://eprints.uthm.edu.my/3600/

This thesis presents positioning system of Automated Guided Vehicles or AGV for short, which is a mobile robot that follows wire or magnetic tape in the floor to navigate from point to another in workspace. AGV serves in industrial fields to convey materials and products around the manufacturing facility or warehouse thus, time of manufacturing process and number of labors can be reduced accordingly. In contrast, the limitation of its movement specified by the guidance path considered as a main weakness. In order to make the AGV moves freely without guidance path, it is essential to know current position first before starts navigate to target place then, the position has to be updating during movement. For mobile robots positioning and path tracking, two basic techniques are usually used, relative and absolute positioning. Relative positioning techniques based on measuring travelled distance by the robot and accumulate it to its initial position to estimate current position, which lead to drift error over time. Digital compass, Global Positioning System (GPS), and landmarks based positioning are examples of absolute positioning techniques, in which robot position estimated from single reading. Absolute positioning does not have drift error but the system cost is high and has signal blockage inside buildings as in case of landmarks and GPS respectively. The developed positioning system based on odometry, accelerometer, and digital compass for path tracking. RFID landmarks installed in predefined positions and ultrasonic GPS used to eliminate drift error in position estimated from odometry and accelerometer. Radio frequency module is used to transfer sensors reading from the mobile robot to a host PC has software program written on LabVIEW, which has a positioning algorithm and graphical display for robot position. The experiments conducted have illustrated that the developed sensor fusion positioning system can be integrated with AGV to replace the ordinary guidance system. It will give AGV flexibility in task manipulation in industrial application.