Abstract:This work presents a 3D multi-robot exploration framework for a team of UGVs moving on uneven terrains. The framework was designed by casting the two-level coordination strategy presented in [1] into the context of multi-robot exploration. The resulting distributed exploration technique minimizes and explicitly manages the occurrence of conflicts and interferences in the robot team. Each robot selects where to scan next by using a receding horizon next-best-view approach [2]. A sampling-based tree is directly expanded on segmented traversable regions of the terrain 3D map to generate the candidate next viewpoints. During the exploration, users can assign locations with higher priorities on-demand to steer the robot exploration toward areas of interest. The proposed framework can be also used to perform coverage tasks in the case a map of the environment is a priori provided as input. An open-source implementation is available online.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how a multi - robot team can perform exploration tasks efficiently and coordinately in complex three - dimensional terrains. Specifically, the researchers proposed a 3D multi - robot exploration framework, which is suitable for a team of unmanned ground vehicles (UGVs) moving on uneven terrains. This framework aims to minimize and explicitly manage conflicts and interferences in the robot team by applying a two - layer coordination strategy to the multi - robot exploration scenario. Each robot uses a prediction - based viewpoint method to select the next scanning position. This method generates candidate next viewpoints by directly expanding the sampling tree in the passable area of the terrain 3D map. In addition, the user can assign positions with higher priorities as needed to guide the exploration direction of the robots. This framework can also be used to perform coverage tasks, especially when the 3D map of the environment is provided as input. The main contributions of the paper are as follows: 1. **Two - layer coordination strategy**: A new two - layer coordination strategy is proposed, which can effectively reduce conflicts and interferences between robots and improve exploration efficiency. 2. **Prediction - based viewpoint selection method**: A prediction - based viewpoint method is developed to select the next best viewpoint by expanding the sampling tree in the passable area of the terrain 3D map. 3. **User priority assignment**: Allows the user to assign positions with higher priorities as needed, thereby guiding the robots to explore specific areas. 4. **Coverage task support**: This framework is not only suitable for exploration tasks but also can be used to perform coverage tasks, especially when the 3D map of the known environment is available. 5. **Open - source implementation**: An open - source implementation is provided, which is convenient for other researchers and developers to use and improve. Through these contributions, this paper provides new solutions and technical support for the autonomous exploration of multi - robot systems in complex environments.

3D Multi-Robot Exploration with a Two-Level Coordination Strategy and Prioritization

Meeting-Merging-Mission: A Multi-robot Coordinate Framework for Large-Scale Communication-Limited Exploration.

3D Multi-Robot Patrolling with a Two-Level Coordination Strategy

Cost-effective Swarm Navigation System Via Close Cooperation

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

Heterogeneous Multi-Robot Collaboration for Coverage Path Planning in Partially Known Dynamic Environments

Autonomous Multi-Robot Exploration Strategies for 3D Environments with Fire Detection Capabilitie

Multi-robot, multi-sensor exploration of multifarious environments with full mission aerial autonomy

Decentralized Strategy for Cooperative Multi-Robot Exploration and Mapping

An Enhanced Hierarchical Planning Framework for Multi-Robot Autonomous Exploration

Cooperative Search and Exploration in Robotic Networks

Graph-based Simultaneous Coverage and Exploration Planning for Fast Multi-robot Search

Multi-Robot Environmental Coverage With a Two-Stage Coordination Strategy via Deep Reinforcement Learning

Coordinated Aerial-Ground Robot Exploration Via Monte-Carlo View Quality Rendering

SPACE: 3D Spatial Co-operation and Exploration Framework for Robust Mapping and Coverage with Multi-Robot Systems

HMS-RRT: A novel hybrid multi-strategy rapidly-exploring random tree algorithm for multi-robot collaborative exploration in unknown environments

Multirobot Unknown Environment Exploration and Obstacle Avoidance Based on a Voronoi Diagram and Reinforcement Learning

A Multi-UAV System for Exploration and Target Finding in Cluttered and GPS-Denied Environments

coExplore: Combining multiple rankings for multi-robot exploration

Aerial-Ground Robots Collaborative 3D Mapping in GNSS-Denied Environments