A 3D Reactive Navigation Algorithm for Mobile Robots by Using Tentacle-Based Sampling
This work addresses navigation in unknown 3D environments for mobile robots, but it is incremental as it extends 2D tentacle-based methods to 3D.
The paper tackles the problem of 3D reactive navigation for mobile robots without global maps by using tentacle-based sampling and heuristic evaluations, achieving fast navigation with statistical results from simulations showing superior performance over a state-of-the-art method.
This paper introduces a reactive navigation framework for mobile robots in 3-dimensional (3D) space. The proposed approach does not rely on the global map information and achieves fast navigation by employing a tentacle based sampling and their heuristic evaluations on-the-fly. This reactive nature of the approach comes from the prior arrangement of navigation points on tentacles (parametric contours) to sample the navigation space. These tentacles are evaluated at each time-step, based on heuristic features such as closeness to the goal, previous tentacle preferences and nearby obstacles in a robot-centered 3D grid. Then, the navigable sampling point on the selected tentacle is passed to a controller for the motion execution. The proposed framework does not only extend its 2D tentacle-based counterparts into 3D, but also introduces offline and online parameters, whose tuning provides versatility and adaptability of the algorithm to work in unknown environments. To demonstrate the superior performance of the proposed algorithm over a state-of-art method, the statistical results from physics-based simulations on various maps are presented. The video of the work is available at https://youtu.be/rrF7wHCz-0M.