Ewok: Real-Time Trajectory Replanning for MAVs using Uniform B-splines and 3D Circular Buffer

For more information see https://vision.in.tum.de/research/robotvision/replanning

1. Related Papers

If you use this code, please cite the respective publications:

• Real-Time Trajectory Replanning for MAVs using Uniform B-splines and 3D Circular Buffer, V. Usenko, L. von Stumberg, A. Pangercic, D. Cremers, In 2017 International Conference on Intelligent Robots and Systems (IROS) [DOI:10.1109/IROS.2017.8202160] [arXiv:1703.01416].

2 Installation

The system has been tested with Ubuntu 20.04 (ROS Noetic) and Ubuntu 18.04 (ROS Melodic). For Ubuntu 16.04 (ROS Kinetic) switch to ubuntu_16_04_kinetic branch.

Follow the tutorials to install ROS Noetic for 20.04 or ROS Melodic for 18.04 (desktop-full) and to set up catkin workspace.

Install dependencies. For Ubuntu 20.04 (ROS Noetic):

sudo apt-get install git python python3-matplotlib python3-numpy libeigen3-dev libgoogle-glog-dev libatlas-base-dev libsuitesparse-dev protobuf-compiler libnlopt-dev libnlopt-cxx-dev ros-noetic-octomap ros-noetic-octomap-ros ros-noetic-octomap-msgs ros-noetic-tf-conversions ros-noetic-eigen-conversions

For Ubuntu 18.04 (ROS Melodic):

sudo apt-get install git python3-matplotlib python3-numpy libeigen3-dev libgoogle-glog-dev libatlas-base-dev libsuitesparse-dev protobuf-compiler libnlopt-dev ros-melodic-octomap ros-melodic-octomap-ros ros-melodic-octomap-msgs ros-melodic-tf-conversions ros-melodic-eigen-conversions

Navigate to the source folder of your catkin workspace, download and build the software:

cd ~/catkin_ws/src
git clone --recursive https://github.com/VladyslavUsenko/ewok.git
cd ..
catkin_make

Some of the provided examples (trajectory_replanning_big_forest.launch) depend on the forest_gen dataset. With the above commands it should already be downloaded as a git submodule. Please check https://github.com/ethz-asl/forest_gen for more information about the dataset.

3. Simulation example

In separate terminal windows:

Start roscore:

roscore

Start visualization:

roscd ewok_simulation/rviz/
rviz -d simulation.rviz

Launch the system and simulator:

roslaunch ewok_simulation trajectory_replanning_simulation.launch

Now you should be able to see Rviz visualization of the system running in simulator. Due to performance reasons GUI of the Gazebo simulator is disabled by default. To enable it, change the following parameter in ewok_simulation/launch/trajectory_replanning_simulation.launch:

<arg name="gui" default="true"/>

4. Circular Buffer vs Octomap benchmarking

In separate terminal windows:

Start roscore:

roscore

or, if you already have it running, make sure you are not using simulated time:

rosparam set use_sim_time false

Start visualization:

roscd ewok_ring_buffer/rviz/
rviz -d ring_buffer.rviz

Run benchmarking script. It will download the dataset, run the computations and show the plots with insertion times:

roscd ewok_ring_buffer/benchmarking/
./benchmark.py

5. B-spline optimization example

In separate terminal windows:

Start roscore:

roscore

or, if you already have it running, make sure you are not using simulated time:

rosparam set use_sim_time false

Start visualization:

roscd ewok_optimization/rviz/
rviz -d optimization_example.rviz

Run optimization example:

rosrun ewok_optimization spline_optimization_example3

6. License

Ewok was developed at the Technical University of Munich. The open-source version is licensed under the GNU Lesser General Public License Version 3 (LGPLv3).