Robot cooperative perception and response is useful in real scenarios, especially in search and rescue scenarios.

Rocopar is developed on Ubuntu 20.04 and ROS Noetic, which is inspired by the following projects:

https://github.com/Icheler/coExplore

https://github.com/MRSD-Team-RoboSAR/robosar

https://github.com/HauserDong/hybrid_control_match/tree/master

• Stage one(now): In ros_stage, leverage a single robot to explore the whole map autonomously, and recognize the target near itself:

  Deadline: May 22

  • Explore the unknown workspace via frontier_exploration
  • Explore the unknown workspace via rrt_exploration
  • Recognize target boxes and publish to the screen.
  • Extend to multi-robot scenarios.

• Stage two: In ros_stage, leverage multiple robots to explore the whole map autonomously, and react to the triggered tasks:

  Deadline: July 30

• Stage three: In gazebo, leverage a fleet of robots to explore the whole map, while recognizing all targets;

  Deadline: Aug 10

• Stage four: In gazebo, explore and response to the triggered tasks.

  Deadline: Sept 10

• Cooperative perception using multi-sensor data fusion
• Reactive planning algorithms for dynamic environments
• Task allocation and coordination among multiple robots
• Integration with ROS for ease of use and extensibility

• ROS Noetic
• Python 3.8+
• Install dependencies: