Online planning for multi-robot active perception with self-organising maps

Abstract

We propose a self-organising map (SOM) algorithm as a solution to a new multi-goal path planning problem for active perception and data collection tasks. We optimise paths for a multi-robot team that aims to maximally observe a set of nodes in the environment. The selected nodes are observed by visiting associated viewpoint regions defined by a sensor model. The key problem characteristics are that the viewpoint regions are overlapping polygonal continuous regions, each node has an observation reward, and the robots are constrained by travel budgets. The SOM algorithm jointly selects and allocates nodes to the robots and finds favourable sequences of sensing locations. The algorithm has a runtime complexity that is polynomial in the number of nodes to be observed and the magnitude of the relative weighting of rewards. We show empirically the runtime is sublinear in the number of

Unsupervised learning‐based flexible framework for surveillance planning with aerial vehicles

“…All the proposed and evaluated algorithms for ‐DTSPN including the SOM‐based method for surveillance planning with Bézier curves can solve problems with a higher number of vehicles; however, we do not consider such scenarios because of the scope of this paper and challenging experimental verification, for example, with 10 vehicles, that needs significantly larger and more demanding experimental setup. Regarding scalability of the used SOM‐based unsupervised learning, it is worth mentioning that it can be considered as independent on the number of vehicles if the number of target locations is significantly higher than the number of vehicles , that is, (see the analysis in Best et al ()).…”

“…Therefore, the complexity grows only with the additional locations that are the individual depots of the vehicles. Hence, the computational complexity can be bounded by which for can be bounded by , and thus it is independent on the number of vehicles (see Best, Faigl, and Fitch () for a detail discussion.…”

“…3 which for ≪ m ncan be bounded by O n ( ) 3 , and thus it is independent on the number of vehicles (see Best, Faigl, and Fitch (2018) for a detail discussion.…”

Collaborative Multi-Robot Search and Rescue: Planning, Coordination, Perception, and Active Vision

“…For instance, in [128], Kurdi et al present a task allocation algorithm for multi-UAV SAR systems inspired by locust insects. Active perception techniques have also been incorporated in multi-robot planning algorithms in existing works [129], [130] An early work in multi-robot task allocation for SAR missions was presented by Hussein et al [122], with a market-based approach formulated as a multiple traveling salesman problem. The authors applied their algorithm to real robots with simulated victim locations that the robots had to divide among themselves and visit.…”

On Realistic Target Coverage by Autonomous Drones

“…In follow-up works, the control strategy was further enhanced to robustly learn and adapt to changes in the environment (Palacios-Gass et al 2016;Schwager et al 2017), while emphasizing decentralized control over a communication networks. More recently, the control strategy was endowed with optimal path planning while avoiding obstacles in the environment (Best et al 2017;Palacios-Gass et al 2017;) and target unpredictability (Hnig and Ayanian 2016). Several challenging aspects of persistent coverage have also been studied: energy-awareness (Derenick et al 2011), connectivity (Orfanus et al 2016), adaptive streaming (Wang et al 2016) and dynamic priorities (da Silva et al 2017).…”