The repository is a benchmark for localizing AttBot2.0 using robot_localization package. The package consists of the following main packages:
-Rosserial: This is used for interfacing with an Arduino Due board which delivers the raw /odom and /imu_data messages as topic to ROS environment.
-Robot_localization: This is the fusion stack that combines the measurements and delivers the filtered odometry in robot_ekf/odom_combined topic.
-hector_slam: for SLAM purposes, and generating a map as well as the /pose in the map.
-RPLidar_ros: this is the driver of the Lidar
An extended kalman filter fuses the odometry and imu data. The IMU data is measured by a Bosch BNO055 sensor while the odometry is the fused result from two motor encoders, a GY-85 (gyro, compass, imu), and an single track model.
The fusion is done in 2D.
A bag data of the measurement is available in the measurements folder.
Start a roscore.
Start the communication with DUE: rosrun rosserial_python serial_node.py _baud:=57600 ...
let the system calibrate, this may take up to 10 seconds and the calibration is over once the red LED flashes.
start the robot_localization stack: roslaunch robot_localization ekf_imu_odometry.launch
start the Lidar stuff:
roslaunch rplidar_ros rplidar.launch
roslaunch hector_slam_launch tutorial_tif.launch