Integration of active localization systems in vehicles control for real-time trajectory tracking

The introduction of "intelligent" mobile vehicles, both in industrial and civil applications, has a slow grow factor due to some fundamental limitations, an obvious one is the limited accuracy in trajectory generation and/or the heavy environmental restructuring required to allow affordable movement accuracy without a direct human supervision. On the other side, active localization systems offer good positioning accuracy, but suffer from timing problems which limit the use in dynamic applications where on-fly trajectory corrections of a pre-defined trajectory are required. In this paper we investigate the problem of integrating the use of active localization systems with traditional odometry in real-time trajectory generation, by the use of nontriangulation based algorithms. Different methods have been studied and tested, both in simulations and real environments. Preliminary experimental results are also presented in the last section of the paper.