Abstract:Autonomous exploration is a fundamental problem for various applications of unmanned aerial vehicles (UAVs). Recently, LiDAR-based exploration has gained significant attention due to its ability to generate high-precision point cloud maps of large-scale environments. While the point clouds are inherently informative for navigation, many existing exploration methods still rely on additional, often expensive, environmental representations. This reliance stems from two main reasons: the need for frontier detection or information gain computation, which typically depends on memory-intensive occupancy grid maps, and the high computational complexity of path planning directly on point clouds, primarily due to costly collision checking. To address these limitations, we present EPIC, a lightweight LiDAR-based UAV exploration framework that directly exploits point cloud data to explore large-scale environments. EPIC introduces a novel observation map derived directly from the quality of point clouds, eliminating the need for global occupancy grid maps while preserving comprehensive exploration capabilities. We also propose an incremental topological graph construction method operating directly on point clouds, enabling real-time path planning in large-scale environments. Leveraging these components, we build a hierarchical planning framework that generates agile and energy-efficient trajectories, achieving significantly reduced memory consumption and computation time compared to most existing methods. Extensive simulations and real-world experiments demonstrate that EPIC achieves faster exploration while significantly reducing memory consumption compared to state-of-the-art methods.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the high memory consumption and computational complexity issues faced during autonomous exploration using unmanned aerial vehicles (UAVs) in large - scale environments. Specifically, many existing exploration methods rely on additional and expensive environmental representations, such as global occupancy grid maps, which lead to high memory requirements in large - scale scenarios. Moreover, the computational complexity of path planning directly on point clouds is relatively high, mainly due to the high cost of collision detection. These problems limit the UAV's effective exploration ability in large - scale environments with limited resources. To address these challenges, the paper proposes EPIC (Exploring on PoInt Clouds), a lightweight LiDAR - based UAV exploration framework that can directly utilize point - cloud data to explore large - scale environments. The main contributions of EPIC include: 1. **A new framework**: This framework directly utilizes point - cloud data for large - scale exploration, eliminating the need for additional expensive environmental representations while ensuring efficient and comprehensive exploration. 2. **Observation graph**: An observation graph directly derived from point - cloud quality, which is used to efficiently guide exploration. 3. **Incremental topological graph construction**: An incremental topological graph construction method that operates directly on point clouds, supports real - time path planning, and promotes efficient exploration in large - scale environments. 4. **Extensive simulation and practical tests**: To verify the superior performance of the proposed method in terms of memory consumption, exploration efficiency, and computational speed. Through these innovations, EPIC can achieve faster large - scale environmental exploration while maintaining low memory consumption and computational time.

EPIC: A Lightweight LiDAR-Based UAV Exploration Framework for Large-Scale Scenarios

REAL: Rapid Exploration with Active Loop-Closing toward Large-Scale 3D Mapping using UAVs

Efficient Autonomous UAV Exploration Framework with Limited FOV Sensors for IoT Applications

Bubble Explorer: Fast UAV Exploration in Large-Scale and Cluttered 3D-Environments using Occlusion-Free Spheres

STExplorer: A Hierarchical Autonomous Exploration Strategy with Spatio-Temporal Awareness for Aerial Robots

RACER: Rapid Collaborative Exploration With a Decentralized Multi-UAV System

Topological Map-Based Autonomous Exploration in Large-Scale Scenes for Unmanned Vehicles

Autonomous Exploration Method for Fast Unknown Environment Mapping by Using UAV Equipped with Limited FOV Sensor

EFP: Efficient Frontier-Based Autonomous UAV Exploration Strategy for Unknown Environments

Autonomous 3D Exploration of Large Structures Using an UAV Equipped with a 2D LIDAR

TDLE: 2-D LiDAR Exploration With Hierarchical Planning Using Regional Division

An autonomous UAV system based on adaptive LiDAR inertial odometry for practical exploration in complex environments

An Enhanced Hierarchical Planning Framework for Multi-Robot Autonomous Exploration

Multi-Drone Collaborative Trajectory Optimization for Large-Scale Aerial 3D Scanning

Multi-map Collaborative Exploration Toward Large-Scale 3D Mapping Using UAV *

An informative path planning framework for UAV-based terrain monitoring

Towards Multi-robot Exploration: A Decentralized Strategy for UAV Forest Exploration

FAEP: Fast Autonomous Exploration Planner for UAV Equipped with Limited FOV Sensor

SEER: Safe Efficient Exploration for Aerial Robots using Learning to Predict Information Gain

Multi-UAVs Collaborative Path Planning in the Cramped Environment

Enhancing Cooperative Exploration and Planning: UAV-Legged Robot Synergy