STAM: a system of tracking and mapping in real environments

Abstract

We have implemented a system of tracking mobile robots and mapping an unstructured environment, using up to 25 wireless sensor nodes in an indoor setting. These sensor nodes form an ad hoc network of beacons, self-localize with respect to three anchor nodes, and then track the locations of mobile robots in the field. The system described here was motivated by search and rescue applications, and has been demonstrated in real physical environments.