Spider-Tree: Ordered Rooted Tree-Based Entrance Organizer for Unknown Area Exploration

Autonomous exploration in complex unknown indoor environments poses challenges for mobile robots, particularly in detecting and navigating through multiple sub-rooms and doorways. Traditional frontier-based methods often struggle with efficient entrance selection and backtracking. This paper proposes the Spider-Tree Algorithm, a hierarchical exploration strategy that models entrances as nodes in an ordered rooted tree. By structuring rooms and sub-rooms as parent-child relationships, the robot systematically explores and revisits all areas. The Spider-Tree algorithm outperforms the Frontier-RRT and Tree-Based methods by reducing the average travel distance by 68.3 % and 77.8 %, and decreasing exploration time by 262 seconds (44.0 %) and 591 seconds (63.9 %) respectively, demonstrating its optimum efficiency in structured, multi-room unknown indoor environments.