Sample-Based Output-Feedback Navigation with Bearing Measurements

TL;DR

This paper introduces a sample-based feedback motion planning method using bearing measurements from landmarks, enabling navigation with monocular cameras and addressing practical challenges like model mismatches and limited field of view.

Contribution

It presents a novel strategy that utilizes bearing-only measurements for feedback control, simplifying sensor requirements compared to previous displacement-based methods.

Findings

Successful simulation validation of the algorithm.

Experimental results demonstrate robustness to model mismatches.

Effective navigation with monocular camera measurements.

Abstract

We consider the problem of sample-based feedback-based motion planning from bearing (direction-only) measurements. We build on our previous work that defines a cell decomposition of the environment using RRT*, and finds an output feedback controller to navigate through each cell toward a goal location using duality, Control Lyapunov and Barrier Functions (CLF, CBF), and Linear Programming. In this paper, we propose a novel strategy that uses relative bearing measurements with respect to a set of landmarks in the environment, as opposed to full relative displacements. The main advantage is then that the measurements can be obtained using a simple monocular camera. We test the proposed algorithm in the simulation, and then in an experimental environment to evaluate the performance of our approach with respect to practical issues such as mismatches in the dynamical model of the robot, and measurements acquired with a camera with a limited field of view.