Multisensor Based Indoor Vehicle Localization System for Production and Logistic

In this paper a multisensor based indoor vehicle localization system for production and logistics is introduced. To track the position and the orientation of a moving vehicle a set of distance values to several points on the vehicle is measured by a wireless ranging system. The beacons of the wireless system are mounted at known positions in the surrounding infrastructure. It is shown that conventional multilateralization is not very practical to solve the required positioning task. In order to match the complete geometry of the forklift to a set of measured distance data, a heuristic nonlinear optimization method is applied. With our novel approach it is possible to solve the complex underlying transformation problem and to calculate the position and angle of the forklift for nearly arbitrary measuring conditions. The achieved accuracy is optimal in the least squares sense. For situations where the wireless access to the vehicle is disturbed, the localization system is assisted by data from a laser scanner. By matching subsequent scans relative movements of the vehicle are be determined precisely. The fusion of an optical relative sensors and a wireless absolute localization system allows for a flexible and steady control of transportation processes even in complex and dynamically changing environments