Kitware SLAM

SLAM and navigation implementation for kitbot

• Install necessary packages for SLAM with LIDAR:
  • ldlidar
  • ydlidar
  • scan matcher
  • ros2 laser scan matcher
  • slam_toolbox
• Use kitware to drive the robot to a pose set by rviz2

Setup

1. Clone into the src directory of your colcon workspace for ROS
2. Make sure you have cloned kitware and kitware_interface in src as well
3. cd in to the kitware_slam and run setup.sh
4. Run colcon build from the workspace folder

The following steps should already be done on the hardware we give students.

1. Make sure you have TAMProxy-Firmware running on a teensy
2. Install TAMProxy-pyHost

Modify for your robot

• Change the transformation for base_link_to_base_laser_tf_node in launch/common.launch.py to reflect the location of the LIDAR's base to the center of your robot. The argument format is

  [ x, y, z, roll, pitch, yaw, parent_frame, child_frame]

  or

  [ x, y, z, qx, qy, qz, qw, parent_frame, child_frame]

• Tuning the PID and error tolerance constants in kitware/differential_driver.py for desired performance/accuracy

Important

Run colcon build from the workspace folder to update the changes after modifying. Else the changes will not be applied

Running

1. Make sure the robot is running off the battery and not attached to anything
2. ssh -X into the robot
3. cd into your ROS workspace
4. Run source install/setup.bash to setup your environment to use the built packages
5. Launch

• If using LDLIDAR: run ros2 launch kitware_slam ldlidar_slam.launch.py
• If using YDLIDAR: run ros2 launch kitware_slam ldlidar_slam.launch.py

6. Open another terminal and ssh -X into the robot again
7. Run rviz2
8. Open rviz configuration :TODO:
9. Set a goal pose
10. Cross your finger that the robot will go

Troubleshoot

• Make sure the kitbot works with keyboard. If not then troubleshoot that first (pin settings in kitbot.py is correct, motor wiring is correct)