Title:Act, Perceive, and Plan in Belief Space for Robot Localization

Abstract:In this paper, we outline an interleaved acting and planning technique to rapidly reduce the uncertainty of the estimated robot's pose by perceiving relevant information from the environment, as recognizing an object or asking someone for a direction.
Generally, existing localization approaches rely on low-level geometric features such as points, lines, and planes, while these approaches provide the desired accuracy, they may require time to converge, especially with incorrect initial guesses. In our approach, a task planner computes a sequence of action and perception tasks to actively obtain relevant information from the robot's perception system. We validate our approach in large state spaces, to show how the approach scales, and in real environments, to show the applicability of our method on real robots.
We prove that our approach is sound, complete, and tractable in practical cases.