Scalable Coverage Path Planning of Multi-Robot Teams for Monitoring Non-Convex Areas

Abstract

This paper presents a novel multi-robot coverage path planning (CPP) algorithm -aka SCoPP -that provides a time-efficient solution, with workload balanced plans for each robot in a multi-robot system, based on their initial states. This algorithm accounts for discontinuities (e.g., nofly zones) in a specified area of interest, and provides an optimized ordered list of way-points per robot using a discrete, computationally efficient, nearest neighbor path planning algorithm. This algorithm involves five main stages, which include the transformation of the user's input as a set of vertices in geographical coordinates, discretization, load-balanced partitioning, auctioning of conflict cells in a discretized space, and a path planning procedure. To evaluate the effectiveness of the primary algorithm, a multi-unmanned aerial vehicle (UAV) post-flood assessment application is considered, and the performance of the algorithm is tested on three test maps of varying sizes. Additionally, our method is compared with a state-of-the-art method created by Guasella et al. Further analyses on scalability and computational time of SCoPP are conducted. The results show that SCoPP is superior in terms of mission completion time; its computing time is found to be under 2 mins for a large map covered by a 150-robot team, thereby demonstrating its computationally scalability.

Faster navigation of semi-structured forest environments using multirotor UAVs

“…Depending on the use case, completion of the map can range from localization of an object to observing the entire traversable volume in the environment. The exploration objective also differs depending on the application; some examples are minimizing exploration time [1,2,10,11], maximizing reconstruction accuracy [3,4], dealing with multirobot and battery limitations [12], object search [13], or multiagent exploration [14,15]. These methods can be broadly classified into sampling-or frontier-based methods.…”

Learning to Coordinate for a Worker-Station Multi-robot System in Planar Coverage Tasks

“…Kapoutsis et al uses an area division algorithm to allocate tasks for multiple robots [11]. Apart from graph-based methods, decomposition-based methods also take large parts in the literature [12], [13], which first partition the target area into obstacle-free convex sub-regions for different robots and then apply single robot coverage planning for each robot separately. Most graphbased or decomposition-based mCPP methods do offline planning, and some also require the coverage area to satisfy specific assumptions (e.g., convex-shaped area).…”

Vision-based Distributed Multi-UAV Collision Avoidance via Deep Reinforcement Learning for Navigation

“…However, the path planning algorithms for UAVs differ from those of UGVs in that the collision avoidance problem needs to be solved in the three-dimensional workspaces, which poses much more difficulties and challenges. Some researchers propose solutions to multi-UAV collision avoidance systems based on centralized methods, which assume that the actions of all UAVs are determined by a central server that can access global information about UAVs (e.g., all UAVs' locations) and workspaces (e.g., a grid map) [7], [8]. However, the centralized multi-UAV systems rely heavily on communication between UAVs and the central server.…”