Mobile Robot Map Learning from Range Data in Dynamic Environments

Abstract

The problem of generating maps with mobile robots has received considerable attention over the past years. Most of the techniques developed so far have been designed for situations in which the environment is static during the mapping process. Dynamic objects, however, can lead to serious errors in the resulting maps such as spurious objects or misalignments due to localization errors. In this chapter, we consider the problem of creating maps with mobile robots in dynamic environments. We present two approaches to deal with non-static objects. The first approach interleaves mapping and localization with a probabilistic technique to identify spurious measurements. Measurements corresponding to dynamic objects are then filtered out during the registration process. Additionally, we present an approach that learns typical configurations of dynamic areas in the environment of a mobile robot. Our approach clusters local grid maps to identify the typical configurations. This knowledge is then used to improve the localization capabilities of a mobile vehicle acting in dynamic environments. In practical experiments carried out with a mobile robot in a typical office environment, we demonstrate the advantages of our approaches.