Running the EKF filter for localization

Table of Contents

• Ndt EKF smooth localization

Ndt EKF smooth localization

This demo aims to show Kalman filter smoothing on top of NDT localization. The pose messages from NDT get a covariance assigned, then get passed into the Kalman filter node. The smooth output is then published as an Odometry topic and visualized in rviz2. This demo is tested with the Lexus car in LG simulator on the parking lot map.

Before running the demo, ensure that the simulator is running:

Start simulation as described in LGSVL simulator. Additionally, to configure LGSVL for this demonstration:

1. Maps: use this map link
2. Vehicles: Select ROS2 native bridge type and paste the content of AutowareAuto/lgsvl-sensors.json into the Sensors text box
3. Simulations: In General tab, Select Cluster = Local Machine and untick any boxes. In Map & Vehicles tab, ensure to untick Run simulation in interactive mode. In Traffic tab, untick all selection. The Weather tab is irrelevant.

Then, run the demo as follows:

$ ade enter
ade$ source /opt/AutowareAuto/setup.bash
ade$ ros2 launch autoware_demos ekf_ndt_smoothing_lgsvl.launch.py

The localization module will automatically locate the vehicle in the parking lot. You can use the controls to move the vehicle in the simulator and observe the red arrow move in tandem. For information about the control refer to Starting the simulation. The following image shows the simulation working as expected: