Throughout all previous chapters of this book, we assumed that the environment does not change over time. This assumption however is not realistic especially for environments populated by humans. People typically walk around, open and close doors, add or remove things, or even move objects like furniture. In the literature, most of the approaches to mapping with mobile robots are based on the assumption that the environment is static. As reported by Wang and Thorpe [158] as well as by Hähnel et al. [61], dynamic objects can lead to serious errors in the resulting map. A popular technique to deal with non-static environments is to identify dynamic objects and to filter out the range measurements reflected by these objects. Such techniques have been demonstrated to be more robust than traditional mapping approaches. They allow a robot, for example, to filter out walking people or passing cars. Their major disadvantage lies in the fact that the resulting maps only contain the static aspects of the environment.