Efficient frontier detection for robot exploration

Abstract

Frontier-based exploration is the most common approach to exploration, a fundamental problem in robotics. In frontierbased exploration, robots explore by repeatedly detecting (and moving towards)

Efficient Dense Frontier Detection for 2-D Graph SLAM Based on Occupancy Grid Submaps

“…However, this approach is not feasible for larger maps and real-time robot operation with such maps, as it presents a significant computational burden. 1) Keidar and Kaminka's seminal work on efficient frontier detection: Keidar and Kaminka [10] proposed in 2014 several approaches which attempt to perform frontier detection in an efficient manner. The first, Wavefront Frontier Detector (WFD), consists of running two consecutive breadth-first searches (BFS).…”

Incremental algorithms for Safe and Reachable Frontier Detection for robot exploration

“…The common approach of all of these frontier detection strategies is to bound the processing of the map to only the updated regions in the map. Two algorithms, Wavefront Frontier Detector (WFD) and Fast Frontier Detector (FFD) are proposed in [15]. WFD is based on the idea of processing only the mapped area instead of the entire map negating the need to unnecessarily process unknown map cells to detect frontiers.…”

“…WFD-INC [15] is a modification to the WFD method by restricting the breadth first search for frontiers only to the active region in the map (i.e. locally updated region).…”

“…This continuous mapping results in a frequent evolution of frontiers in the occupancy grid map. It is suggested in the literature that making use of these frequently evolving frontier information could lead to more efficient exploration strategies [14][15][16]. Having access to frequently evolving frontiers allow robots to quickly uncover dead-end situations and traps.…”

“…Having access to frequently evolving frontiers allow robots to quickly uncover dead-end situations and traps. It also allows a robot to discover when the target location assigned is fully or substantially explored by another robot due to unintentional crossing of paths during multi-robot missions [15]. These early discoveries of undesirable situations during exploration missions reduce unnecessary motions for robots thus improves the overall efficiency of missions [17].…”