doomse / frontier_exploration

A frontier exploration module implementied with ROS 2, C++, and Python.

frontier_exploration

Description

A frontier exploration module implementied with ROS 2, C++, and Python. Based on:

B. Yamauchi, "A frontier-based approach for autonomous exploration," Proceedings 1997 IEEE International Symposium on Computational Intelligence in Robotics and Automation CIRA'97. 'Towards New Computational Principles for Robotics and Automation', Monterey, CA, USA, 1997, pp. 146-151, doi: 10.1109/CIRA.1997.613851.

Frontier Exploration: Contains nodes for autonomous exploration, interfaces with the nav2 stack.

Frontier Interfaces: Contains a custom ROS 2 service interface.

Learned frontier detector: Contains scripts and tools for training and deploying a learned frontier detector.

Getting Started

To get ready for development or execution:

# Clone the repo
git clone https://github.com/adrian-soch/frontier_exploration

# Build
colcon build

To run the exploration node:

# Start rosbag or robot simulation or real robot

ros2 launch frontier_exploration exploration.launch.py

# The node can be run independantly via ros2 run
# And can be tested with:
ros2 service call  /frontier_pose frontier_interfaces/srv/FrontierGoal goal_rank:\ 0
#requester: making request: frontier_interfaces.srv.FrontierGoal_Request(goal_rank=0)

Output can be visualized in rviz with the frontier map and frontier region markers.

To run the full exploration system:

# Start rosbag or robot simulation or real robot
ros2 launch frontier_exploration lite_turtlebot_full_stack.launch.py

# Launch the frontier exploration node
ros2 launch frontier_exploration exploration.launch.py

Training Data Collection

1. Launch a simalation or real robot that publishes occupancy grid maps to the /map topic.
2. Then ros2 run frontier_exploration collection_node
   • This will preprocess the map through simple filtering operations
   • Save a map after the robot has travelled beyond a distance threshold

Dataset

Our dataset is available at the link above.