Frontier-Based Exploration
Overview
Frontier-based exploration studies how robots can choose where to move next by identifying boundaries between known and unknown space. Common methods rank frontier viewpoints using information gain, travel cost, terrain and energy constraints, semantic cues, or task goals, then combine these scores with mapping and path planning. Work spans indoor and outdoor robots, drones, quadrupeds, and multi-robot teams, with evaluations in simulated environments and real-world exploration and navigation tasks. A central challenge is balancing coverage and task success against time, energy, safety, and computation while handling uncertain predictions, difficult terrain, and changing communication or mission constraints.
Latest papers
28 papers1 paper in the last four weeks, with none the four weeks before. 0.0% of all new papers.