Development of an automated guided vehicle for use in industrial applications

Abstract

This paper presents a navigation system for an Automated Guided Vehicle (AGV) based on existing IEEE 802.11 wireless infrastructure in place for localization and navigation, through the use of multiple wireless access points. A fingerprinting based method is used to deduce the location based on the wireless signal intensities and a Kalman filter based approach is used to estimate the location of the unit accurately. The proposed method is able to break free of the accumulation of error in pure odometry based system and the error in pure signal intensity based systems due to variability of the received signal. The system incorporates the localization system with a navigation system including obstacle detection and avoidance to improve upon existing AVG navigation systems.