This tab shows a 3D visualization of the robot and field. It can be used with regular 2D odometry, but is especially helpful when working with 3D calculations (like localizing with AprilTags). Multiple camera views are available, including field relative, robot relative, and fixed. The timeline shows when the robot is enabled and can be used to navigate through the log data.
Note: To view the 3D visualization alongside other tabs, click the "Add Window" icon just below the navigation/playback controls. Each pop-up window can also be configured with a different camera view (see more details below). To hide the controls at the bottom of the window, click the eye icon.
To add a field with pose data, drag it from the sidebar to the box under "3D Poses" or "2D Poses" and use the drop down to select an object type. Multiple sets of objects can be added this way, and fields can be included multiple times. To remove a set of objects, right-click the field name. The origin and units are configurable.
Pose data can be stored as a byte-encoded struct or protobuf. The following data types are supported:
Pose3dPose2dTranslation3dTranslation2dTransform3dTransform2dAprilTagTrajectory
The example code below shows how to log pose data using WPILib or AdvantageKit.
Pose3d poseA = new Pose3d();
Pose3d poseB = new Pose3d();
// WPILib
StructPublisher<Pose3d> publisher = NetworkTableInstance.getDefault()
.getStructTopic("MyPose", Pose3d.struct).publish();
StructArrayPublisher<Pose3d> arrayPublisher = NetworkTableInstance.getDefault()
.getStructArrayTopic("MyPoseArray", Pose3d.struct).publish();
periodic() {
publisher.set(poseA);
arrayPublisher.set(new Pose3d[] {poseA, poseB});
}
// AdvantageKit
Logger.recordOutput("MyPose", poseA);
Logger.recordOutput("MyPoseArray", poseA, poseB);
Logger.recordOutput("MyPoseArray", new Pose3d[] {poseA, poseB});WPILib's Field2d class can also be used to log several sets of 2D pose data together. Note that Field2d publishes rotations in degrees instead of radians; use the configuration at the bottom of the screen to adjust the units.
Alternatively, pose data can be stored as a numeric array describing one or more poses with the formats shown below.
2D Poses
[
x, y, rot,
x, y, rot,
...
]
The 2D rotation must be CCW+, and the units (radians/degrees) are configurable.
3D Poses
[
x, y, z, w_rot, x_rot, y_rot, z_rot,
x, y, z, w_rot, x_rot, y_rot, z_rot,
...
]
The w, x, y, and z rotation values represent a quaternion, which is used internally by WPILib's Rotation3d class.
The following objects are supported:
- Robot
- Ghost (Green, Yellow, Blue, & Red)
- AprilTag (36h11 & 16h5)
- Axes
- Trajectory
- Vision Target
- Game Pieces
- Blue Cone (Front/Center/Back)
- Yellow Cone (Front/Center/Back)
- Camera Override (more details under "Camera Options")
- Mechanism (more details under "Mechanisms & Components")
- Component (more details under "Mechanisms & Components")
Note: AprilTags use a smile texture by default. Add a field as "AprilTag ID" to specific an ID (0-586 for 36h11, 0-29 for 16h5) for each displayed tag. The ID field must be a numeric array where each item corresponds to a tag.
Mechanism data can be visualized using 2D mechanisms or articulated 3D components.
To visualize mechanism data logged using a Mechanism2d, drag the mechanism field to "2D Poses". The mechanism is projected onto the XZ plane of the robot using simple boxes (as shown below). The robot's origin is centered on the bottom edge of the mechanism.
Multiple mechanisms can be displayed simultaneously if more than one mechanism field is provided. Mechanisms are displayed on the main (solid) robot by default, but can also moved to the ghost robot by switching from "Mechanism (Robot)" to "Mechanism (Ghost)".
To visualize mechanism data using articulated 3D components, log a set of 3D poses with the robot-relative locations of each component. Each component can be moved independently (like an elevator carriage, arm, or end effector). Drag the component data to "3D Poses" and select "Component (Robot)" or "Component (Ghost)". Component data can be provided separately for the ghost robot (for example, to visualize a mechanism setpoint). For more information on configuring robots with components, see Custom Assets.
Each field includes a set of game piece object types, allowing game pieces to be rendered at any position on the field using data published by the robot code. This has a variety of applications, including:
- Visualizing the actions of simulated auto routines using simple animations
- Showing the detected locations of game pieces on the field
- Indicating where game pieces are located within a robot
- Viewing shot trajectories based on physics calculations
The AdvantageKit KitBot 2024 example project includes a simple example of a command that animates a note traveling from the robot to the speaker. This command is incorporated into the standard launch sequence, triggering the animation whenever a note is released. This video shows how game piece animations can be used to visualize auto routines for several different games.
Another simple use case is showing the state of game pieces within the robot based on sensor data. For example, a beam break sensor within the note path for a 2024 robot could cause a note to appear:
To switch the selected camera mode, right-click on the rendered field view. The camera mode and position is controlled independently for every pop-up window, allowing for the easy creation of multi-camera views.
Note: Right-click the rendered field view and click "Set FOV..." to adjust the FOV of the orbiting and Driver Station cameras.
This is the default camera mode, where the camera can be freely moved relative to the field. Left-click + drag rotates the camera, and right-click + drag pans the camera. Scroll to zoom in and out.
This mode has the same controls as the "Orbit Field" mode, but the camera's position is locked relative to the robot. This allows for "tracking" shots of the robot's movement.
This mode locks the camera behind one of the driver stations at typical eye-height. Either manually choose the station to view or choose "Auto" to use the alliance and station number stored in the log data.
Note: Automatic selection of station number may be inaccurate when viewing log data produced by AdvantageKit 2023 or earlier.
Each robot model is configured with a set of fixed cameras, like vision and driver cameras. These cameras have fixed positions, aspect ratios, and FOVs. These views are often useful to check vision data or to simulate a driver camera view. In the example below, a driver camera is shown.
If a "Camera Override" pose is provided, it replaces the default poses of all fixed cameras while retaining their configured FOVs and aspect ratios. This allows the robot code to provide the position of a moving camera, like one mounted to a turret or shooter hood.
Note: As with all other pose data, the "Camera Override" pose must be field relative, not robot relative.
The following configuration options are available:
- Field: The field model to use, defaults to the most recent game. We recommend using the "Evergreen" field for devices with limited graphical performance. The "Axes" field displays only XYZ axes at the origin with a field outline for scale.
- Alliance: The position of the field origin, on the blue or red alliance wall. "Auto" will select the alliance color based on the available log data.
- Robot: The robot model to use. We recommend using the "KitBot" model for devices with limited graphical performance. To add a custom field or robot model, see Custom Assets.
- Units: The linear and angular units of the provided fields. Meters, inches, radians, and degrees are supported. The rotations units do no affect 3D poses.
Note: Automatic selection of alliance color may be inaccurate when viewing log data produced by AdvantageKit 2023 or earlier.
The 3D field supports three rendering modes:
- Cinematic: Render using shadows, lighting, and reflections for a more realistic look. Requires a decently powerful GPU.
- Standard (Default): Render with minimal lighting (no functional difference from cinematic mode).
- Low Power: Lower the framerate and resolution to reduce battery consumption and provide more consistent performance on low-end devices.
To configure the rendering mode, open the preferences window by pressing cmd/ctrl + comma or clicking "Help" > "Show Preferences..." (Windows/Linux) or "AdvantageScope" > "Settings..." (macOS). The "3D Mode (Battery)" setting can be switched from the default to override the rendering mode used on a laptop when not charging. For example, this can be used to preserve battery while at competition.
