Simultaneous Localization and Planning for Physical Mobile Robots via Enabling Dynamic Replanning in Belief Space

Simultaneous planning while localizing is a crucial ability for an autonomous robot operating under uncertainty. This paper addresses this problem by designing methods to dynamically replan while the localization uncertainty or environment map is updated. In particular, relying on sampling-based methods, the proposed online planning scheme can cope with challenging situations, including when the localization update or changes in the environment alter the homotopy class of trajectories, in which the optimal plan resides. The proposed algorithm eliminates the need for stabilization in the state-of-the-art FIRM (Feedback-based Information RoadMap) method, and outperforms its performance and success probability. Applying belief space planning to physical systems brings with it a plethora of challenges. Thus, designing computationally tractable algorithms, a key focus and contribution of this paper is to implement the proposed planner on a physical robot and show the SLAP (simultaneous localization and planning) performance under uncertainty, in changing environments, and in the presence of large disturbances such as kidnapped robot situation.