The Particle Filter in Practice

Positioning of moving platforms has been a technical driver for real-time applications of the particle filter (PF) in both the signal processing and the robotics communities. For this reason, we will spend some time to explain several such applications in detail, and to summarize the experiences of using the PF in practice. The applications concern positioning of underwater vessels, surface ships, cars, and aircraft using geographical information systems containing a database with features of the surrounding. In the robotics community, the PF has been developed into one of the main algorithms (FastSLAM) for solving the simultaneous localization and mapping (SLAM) problem. This can be seen as an extension to the aforementioned applications, where the features in the geographical information system are dynamically detected and updated on the fly. The common denominator of these applications of the PF is the use of a low-dimensional state vector consisting of horizontal position and course (three dimensional pose). The PF performs quite well in a three dimensional statespace. However, the PF is not practically useful when extending the models to more realistic cases with