Dynamic Frontier Based Exploration with a Mobile Indoor Robot

Any indoor service robot needs a kind of map as knowledge of its working space in order to perform navigation tasks efficiently. Ideally the robot should be able to generate this map autonomously when it is placed in a priori unknown environment. Thus it has to explore its surroundings completely in a reasonable amount of time and produce a map as accurate and complete as possible. This paper presents an integrated dynamic exploration approach that combines existing solutions of mapping, localization, obstacle avoidance and navigation to realize an efficient exploration strategy.