Concurrent localisation and map building for mobile robots using ultrasonic sensors

Successful autonomous navigation by a mobile robot depends on its ability to map its environment and to accurately determine its position within this environment. In complex environments, where no additional navigational aids are present, the localization and map-building processes are not independent of each other. This paper presents an approach where the boot-strapping problem of concurrent localization and map building is solved by estimating the respective errors introduced by each of the processes and correcting them accordingly. The algorithms were tested in simulation, and results showed that the robot stably navigated in its environment.