Development of an autonomous guided vehicle for indoor propagation measurements

Abstract

At Virginia Tech, an autonomous guided vehicle (AGV) has been developed for use in indoor propagation measurements and for a research/teaching tool in the Computer Integrated Manufacturing Laboratory. The AGV design strategy, navigation methodology, and a calibration technique used to minimize path errors caused by unequal wheel radii in dead-reckoning navigation are described. The basic geometry of dead-reckoning navigation is developed, with a particular emphasis on the use of a calibration run to yield exact wheel radii and steering correction factors for use in navigation. These techniques were used on an actual AGV, and measurements were taken under operating conditions to determine navigation accuracy. The major limitation in dead reckoning comes from improper estimates of wheel radii. This can be compensated for with a calibration run, but the tire radius may still change due to dynamic conditions such as redistribution of vehicle weight during cornering, and inexact manufacturing of the tires themselves. Measurements of vehicle position through use of calibrated dead reckoning show accurate vehicle position to within a few cm along a 10 m path.<>