Hey everyone, welcome to our project! This is built off of the CURLY DRIFT library, please refer to their wonderful and extremely thorough installation and build instructions below.
In the automotive industry, Inertial Measurement Units (IMU) and Global Positioning System (GPS) sensors are frequently used to obtain vehicle state data. It is common practice to increase sensor accuracy and sensor message frequency to improve algorithm accuracy; however, this usually comes at the expense of a more expensive sensor. This paper aims to investigate the impact of decreasing GPS frequency on an Invariant Kalman filter implementation, in order to comprehend a potential lower bound on GPS sensor update rate with respect to accumulated error.
All project materials can be found in the following folder: https://drive.google.com/file/d/1bfKoXbWuWJHtk1LLLd8McSnwwiFw3PpV/view?usp=drive_link
Dead Reckoning In Field Time (DRIFT) is an open-source C++ software library designed to provide accurate and high-frequency proprioceptive state estimation for a variety of mobile robot architectures. By default, DRIFT supports legged robots, differential-drive wheeled robots, full-size vehicles with shaft encoders and marine robots with a Doppler Velocity Log (DVL). Leveraging symmetry-preserving filters such as Invariant Kalman Filtering (InEKF), this modular library empowers roboticists and engineers with a robust and adaptable tool to estimate instantaneous local pose and velocity in diverse environments. The software is structured in a modular fashion, allowing users to define their own sensor types, and propagation and correction methods, offering a high degree of customization.
Detailed documentations and tutorials can be found at https://umich-curly.github.io/DRIFT_Website/.
We tested our filter implementation on a several GPS frequencies and driving scenarios. The datasets come from the [Ford AV]https://github.com/Ford/AVData dataset. More information on these samples were obtained can be seen in the data README
We have tested the library in Ubuntu 20.04 and 22.04, but it should be easy to compile in other platforms.
We use the threading functionalities of C++17.
Required by header files. Download and install instructions can be found at: http://eigen.tuxfamily.org. Requires at least 3.1.0.
Required by header files. Download and install instructions can be found at: https://github.com/jbeder/yaml-cpp.
Required for trajectory visualization and error calculation. Download and install instructions can be found at: https://github.com/MichaelGrupp/evo
Building with ROS1 is optional. Instructions are found below.
Clone the repository:
git clone https://github.com/abcarr/eecs568-drift-FordAV.git
Create another directory which we will name 'build' and use cmake and make to compile an build project:
mkdir build
cd build
cmake ..
make -j4
After building the library, you can install the library to the system. This will allow other projects to find the library without needing to specify the path to the library.
sudo make install
Then, you can include the library in your project by adding the following line to your CMakeLists.txt file:
find_package(drift REQUIRED)
We provide several examples in the ROS/examples directory.
- Add
/ROS/driftto theROS_PACKAGE_PATHenvironment variable. Open your ~/.bashrc file in a text editor and add the following line to the end. Replace PATH/TO with the directory path to where you cloned drift:
export ROS_PACKAGE_PATH=${ROS_PACKAGE_PATH}:PATH/TO/eecs568-drift-FordAV/ROS/drift
Then
source ~/.bashrc
- Execute
build_ros.shscript in the repository root directory:
cd eecs568-drift-FordAV
chmod +x build_ros.sh
./build_ros.sh
- start roscore
roscore
- in another terminal start ford_av node
rosrun drift ford_av
- in another terminal start rosbag
rosbag play data/FordAV_straight_path_data_sample.bag --pause
- in another terminal start gps throttling (Use 180 Hz as default)
rosrun topic_tools throttle messages gps <desired frequency>
- in another terminal (visualize the pose). Return to terminal with rosbag and press spacebar to "play" demo
rostopic echo /ford_av/vanila_inekf/pose
If you are interested in recording your own bag file for analysis...
rosbag record -a -O example_bagname
To visualize trajectories/error metrics, we used the evo library (linked above). Some example commands # are below, but please refer to the library documentation for more information. Note that /pose_raw refers to the fused GNSS + IMU solution and /pose_localized refers to the Lidar Odometry solution. For more information please refer to the Ford AV dataset documentation
evo_traj bag example_bagname.bag /pose_raw /pose_localized /ford_av/vanila_inekf/pose --ref=/pose_ground_truth --plot
DRIFT is released under a BSD 3-Clause License.