Camera & Image
Camera and image messages are rendered in the Image panel. When a recording contains multiple image topics, Avala displays them in a multi-camera grid layout.
3D Point Clouds & LiDAR
Point cloud messages are rendered in the 3D / Point Cloud panel.
PointCloud2 supports any combination of fields — Avala reads the field descriptors from the message to determine how to interpret the data.
Laser Scan
LaserScan messages are routed to the 3D panel when the topic name contains a recognized keyword (
lidar, pointcloud, velodyne, or points). Topics with unrecognized names fall back to the Raw Messages panel.
Radar
RadarScan messages are routed to the 3D panel when the topic name contains a recognized keyword. Topics with unrecognized names fall back to the Raw Messages panel.
Transforms & Localization
Transform messages are used internally to resolve coordinate frames between sensors. Including
/tf and /tf_static topics in your recordings enables Avala to project LiDAR data onto camera images and align data from sensors mounted at different positions.
NavSatFix messages are rendered in the Map panel, showing the vehicle or robot path on a geographic map.
IMU
IMU messages are rendered in the Plot panel, showing time-series charts of orientation, angular velocity, and linear acceleration components.
Auto-Detection Logic
Avala uses a two-step process to assign each topic to a panel type:-
Schema name matching (primary) — The message schema name is checked against the known types listed above. For example, any topic with schema
sensor_msgs/PointCloud2is assigned to a 3D panel regardless of its topic name. - Keyword matching (fallback) — If the schema name is not one of the built-in types above, Avala checks for keywords in the schema name and topic name:
Schema name matching (step 1) always takes precedence over keyword matching.