Improved Runtime for the Synchronous Multi-door Filling

Abstract

In this paper, a particular type of dispersion is further investigated, which is called Filling. In this problem, robots are injected one by one into an a priori not known area and have to travel across until the whole area is covered. The coverage is achieved by a robotic team whose hardware capabilities are restricted in order to maintain low production costs. This includes limited viewing and communication ranges. In this work, we present an algorithm solving the synchronous Filling problem in O((k + ∆)·n) time steps by n robots with a viewing range of 1 hop, where k is the number of doors, n is the number of vertices of the graph, and ∆ is the maximum degree of the graph. This improves the best previously known running time bound of O(k · ∆ · n). Furthermore, we remove the constraint from the previous algorithm that the door vertices need to have a degree of 1.

Stigmergy-based, Dual-Layer Coverage of Unknown Indoor Regions

“…Conversely, in some works the initial location of the agents may be inside the region [9,15]. The region to be covered may be a continuous part of the two-dimensional plane [25,15,20] or modelled as a discrete graph (usually as a grid graph) [17,7,8,9,13,14]. The region is sometimes restricted to be simply-connected, which means no obstacle is completely surrounded by empty cells [2].…”

“…Additionally, unlike [17] and most works in this area, we focus on external detection of coverage completion rather than just achieving coverage. Hsiang et al's problem setting and model are extended in numerous works including [7,8,9,13,14,3,2]. The closest work to ours among these is [3], in which the concept of covering an environment with two layers of agents is similarly explored.…”

“…For the workers to safely and efficiently navigate the scene, it is desirable to continuously monitor it prior to and during the workers' operations. Several works have proposed addressing this problem by flooding the region with a swarm of simple, airborne agents prior to workers' entrance [15,17,7,14]. The robots' purpose is to explore and monitor the scene in a distributed fashion, providing rescue workers with real time, potentially lifesaving information.…”