Online Decentralized Perception-Aware Path Planning for Multi-Robot Systems

Abstract

This paper proposes a decentralized and online optimal perception-aware strategy for multi-robot systems. The aim is to maximize the information collected along the planned trajectory about the relative configurations of the robots and, hence, to minimize the localization uncertainty. This is done by leveraging the so-called Constructability Gramian (CG), which can quantify the information about the future state of a nonlinear system. We show that, thanks to a proper change of coordinates, the CG can be computed in a decentralized way with only minor approximations. This allows for formulating an online and decentralized trajectory generation problem for optimal localization. To show the effectiveness of the approach, we consider as case study the localization of a quadrotor group with noisy distance measurements and sensing constraints. The results show the interest of the proposed approach.

A comprehensive review of the latest path planning developments for multi-robot formation systems

“…However, in order to minimize the uncertainty of localization, the authors in ref. [143] proposed an optimization path planning algorithm called the online optimization perceptual strategy. On the other hand, different decentralized decision path planning techniques have been applied to UGV systems, including optimization and machine learning [90][91][92], potential field [93][94][95][96][97][98], cellular automaton (CA) method [99,100], hybrid algorithms [101,102], sampling-based [103], and other techniques [104][105][106].…”

Multi-Robot Active Sensing for Bearing Formations

“…In [13], the authors considered the active information gathering for ensuring observability of a multi-robot system with two anchors and range measurements. In our previous work [14], we considered a formation with relative distance measurements, and we maximized the observability of the system over a future prediction horizon. The proposed method could be made distributed but at the cost of being suboptimal w.r.t.…”

Perception-Aware Trajectory Planning for a Pair of Unicycle-Like Robots With Absolute and Relative Ranging Measurements

“…With the same rationale, Salaris et al [7] propose a perception-aware control strategy maximising for local constructibility (see also [8]). These ideas are extended to multiagent systems in [9], where the vehicles use mutual range measurements to localise themselves in the environment. Range-based information have been used in [10], [11] to maximise observability (measured by the Fisher Information Matrix and by the Observability Matrix, respectively), while Mandić et al [12] combine observability maximisation with the leaderfollower idea.…”