Skip to content

Combines the features of DRL-based as well as model based planning approaches by deciding which planner to use depending on current situation.

License

Notifications You must be signed in to change notification settings

Arena-Rosnav/all_in_one

Repository files navigation

all_in_one

This repository contains the results of work presented in All-in-One: A DRL-based Control Switch Combining State-of-the-art Navigation Planners. The All-In-One Planner (AIO) chooses from a given list of local planners in each iteration with the objective to combine the strengths of learning-based planners and classical planners like TEB / MPC.

This code is intended for usage with Arena-Rosnav infrastructure. For more information regarding AIO please refer to the original repository.

Currently we offer a control policy for 4 different robot models:

  • Turtlebot3 Burger
  • Jackal
  • Robotino
  • Youbot

Installation

Activate poetry shell

cd ~/catkin_ws/src/arena-rosnav # Navigate to your arena-bench location
poetry shell

Make sure to source the workspace environment

cd ../.. # navigate to the catkin_ws directory
catkin_make
source devel/setup.zsh # if you use bash: source devel/setup.bash 

Usage

This planner can be chosen using the local_planner argument like so:

roslaunch arena_bringup start_arena.launch local_planner:=aio # Make sure that your virtual env/poetry is activated

For more details regarding usage, please refer to our documentation

About

Combines the features of DRL-based as well as model based planning approaches by deciding which planner to use depending on current situation.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published