ROS 2 Message Types Used by Waypoint

Use this page when a guide step mentions a message field you do not recognize. The official message definitions are still the source of truth, but this page explains the parts this robot actually uses.

Reading a Message

Use this method every time:

Run ros2 interface show MESSAGE_TYPE.
Find nested messages and arrays.
Identify the fields your node must set or read.
Check units.
Set headers and frames when they exist.
Compare the result with ros2 topic echo.

`geometry_msgs/msg/Twist`

Twist
+-- linear: Vector3
|   +-- x: float64
|   +-- y: float64
|   +-- z: float64
+-- angular: Vector3
    +-- x: float64
    +-- y: float64
    +-- z: float64

Waypoint uses linear.x in meters per second and angular.z in radians per second.

message = Twist()
message.linear.x = 0.10
message.angular.z = 0.30
publisher.publish(message)

The other fields stay zero because this is a ground robot that does not strafe, fly, roll, or pitch.

`std_msgs/msg/Header`

Header
+-- stamp
|   +-- sec
|   +-- nanosec
+-- frame_id: string

stamp says when the data applies. frame_id says which coordinate frame the data is expressed in.

message.header.stamp = self.get_clock().now().to_msg()
message.header.frame_id = 'imu_link'

Let ROS fill sec and nanosec from the clock. Do not fill those two fields by hand unless you are writing a special time tool.

`geometry_msgs/msg/Quaternion`

ROS uses 3D orientations even for a flat ground robot. For yaw-only motion:

half_yaw = yaw / 2.0
qx = 0.0
qy = 0.0
qz = math.sin(half_yaw)
qw = math.cos(half_yaw)

(0, 0, 0, 1) means no rotation. (0, 0, 0, 0) is not a valid orientation.

`sensor_msgs/msg/Imu`

Imu
+-- header: Header
+-- orientation: Quaternion
+-- orientation_covariance: float64[9]
+-- angular_velocity: Vector3
+-- angular_velocity_covariance: float64[9]
+-- linear_acceleration: Vector3
+-- linear_acceleration_covariance: float64[9]

Field	MPU6050 provides it?	Required handling
timestamp	no	use the ROS clock
frame	no	use the configured IMU frame
orientation	no	do not fabricate it
angular velocity	yes	scale and convert to rad/s
linear acceleration	yes	scale and convert to m/s^2
covariance	not directly	use honest documented values

Minimum population pattern:

message = Imu()
message.header.stamp = self.get_clock().now().to_msg()
message.header.frame_id = 'imu_link'
message.orientation.w = 1.0
message.orientation_covariance[0] = -1.0
message.angular_velocity.x = gyro_x_rad_s
message.angular_velocity.y = gyro_y_rad_s
message.angular_velocity.z = gyro_z_rad_s
message.linear_acceleration.x = accel_x_mps2
message.linear_acceleration.y = accel_y_mps2
message.linear_acceleration.z = accel_z_mps2

The orientation_covariance[0] = -1.0 marker says orientation is unavailable. Setting orientation.w = 1.0 avoids an invalid all-zero quaternion, but it does not claim the IMU knows the robot’s orientation because the covariance marker already says it does not.

`nav_msgs/msg/Odometry`

Odometry
+-- header: Header
+-- child_frame_id: string
+-- pose: PoseWithCovariance
|   +-- pose: Pose
|   |   +-- position: Point
|   |   +-- orientation: Quaternion
|   +-- covariance: float64[36]
+-- twist: TwistWithCovariance
    +-- twist: Twist
    +-- covariance: float64[36]

message.pose.pose is not a typo. The outer pose contains a pose and covariance. The inner pose is the actual position and orientation. message.twist.twist follows the same pattern.

For this package:

message.header.frame_id = 'odom'
message.child_frame_id = 'base_link'
pose is expressed in odom
velocity is associated with base_link

odom = Odometry()
odom.header.stamp = stamp
odom.header.frame_id = 'odom'
odom.child_frame_id = 'base_link'
odom.pose.pose.position.x = x
odom.pose.pose.position.y = y
odom.pose.pose.orientation.z = qz
odom.pose.pose.orientation.w = qw
odom.twist.twist.linear.x = linear_mps
odom.twist.twist.angular.z = angular_rad_s

`nav_msgs/msg/Path`

Path
+-- header: Header
+-- poses: list of PoseStamped

poses is a Python sequence. Append poses and trim it so RViz does not grow forever.

pose = PoseStamped()
pose.header = odom.header
pose.pose = odom.pose.pose
self.path.header.stamp = odom.header.stamp
self.path.header.frame_id = 'odom'
self.path.poses.append(pose)
self.path.poses = self.path.poses[-self.path_max_poses:]

`geometry_msgs/msg/TransformStamped`

TransformStamped
+-- header: Header
|   +-- stamp
|   +-- frame_id
+-- child_frame_id: string
+-- transform
    +-- translation: Vector3
    +-- rotation: Quaternion

A transform connects two frames. For odometry:

transform = TransformStamped()
transform.header.stamp = stamp
transform.header.frame_id = 'odom'
transform.child_frame_id = 'base_link'
transform.transform.translation.x = x
transform.transform.translation.y = y
transform.transform.rotation.z = qz
transform.transform.rotation.w = qw
self.tf_broadcaster.sendTransform(transform)

For a fixed IMU mounting, publish or describe base_link -> imu_link. The transform must match the physical mounting, and you should not also rotate the raw IMU axes the same way in code.

Parameter YAML Shape

node_name:
  ros__parameters:
    parameter_name: value

The outer node_name must match the node name in the launch file. YAML indentation matters: use spaces, not tabs.