Skip to main content
Avala natively supports MCAP along with ROS 1 and ROS 2 bag formats. Upload recordings from your robot fleet and visualize camera images, LiDAR point clouds, radar scans, and more — all time-synchronized in a configurable multi-panel viewer. Then annotate the same data without switching tools.

​
Supported Container Formats

FormatExtensionNotes
MCAP.mcapRecommended. Supports all serialization formats with the lowest overhead.
ROS 1 Bag.bagAutomatically converted to MCAP on upload.
ROS 2 Bag.db3 / folderSQLite-based storage. Automatically converted to MCAP on upload.
Use MCAP as your recording format whenever possible. It has the smallest container overhead, supports every serialization framework (Protobuf, JSON, ROS, Flatbuffers), and is the only format that preserves all schema metadata without conversion.

​
What Avala Does With Your MCAP Files

1

Parse and discover

Avala reads the file and discovers every topic and its associated schema. No external dependencies or separate schema files are needed when schemas are embedded in the recording.
2

Auto-detect panel types

Each topic is matched to the best panel type based on its schema name and topic name. For example, a sensor_msgs/PointCloud2 topic is assigned to a 3D point cloud panel, while a sensor_msgs/CompressedImage topic gets an image panel.
3

Build a multi-panel layout

Avala constructs a configurable multi-panel layout so you can view all sensor streams at once. Panels can be resized, rearranged, added, or removed.
4

Synchronize by timestamp

All topics are indexed by message timestamp so playback across panels stays in sync. Scrub the timeline and every panel updates to the same moment in time.
5

Enable annotation

Once the recording is indexed, the same viewer becomes an annotation workspace. Place 3D cuboids on point clouds, draw bounding boxes on camera images, and track objects across frames — all within one interface.

​
Upload Workflow

  1. Navigate to Datasets and create a new dataset (or select an existing one).
  2. Click Upload and select your .mcap, .bag, or .db3 files.
  3. Avala parses the file, discovers topics, and presents a topic selection screen.
  4. Choose the topics you want to visualize and annotate (e.g., /camera/front/image_raw, /lidar/points).
  5. Avala indexes the selected topics by timestamp and creates annotatable frames.
For large files, you can also upload via the API: 
curl -X POST https://api.avala.ai/api/v1/datasets/{dataset_id}/upload \
  -H "X-Avala-Api-Key: your-api-key" \
  -F "file=@recording.mcap"
MCAP files are processed server-side during upload. Large files may take several minutes to index. You will receive a notification when processing is complete.

​
Key Features

Multi-Sensor Playback

View camera feeds, point clouds, radar, IMU, and GPS data simultaneously in a time-synchronized multi-panel viewer.

Auto-Detection

Topics are automatically matched to the correct panel type based on schema and topic name. No manual configuration required.

Configurable Layout

Resize, rearrange, add, and remove panels to build the viewer layout that fits your workflow.

Annotation

Annotate directly on the visualized data. Place 3D cuboids, draw bounding boxes, and track objects without leaving the viewer.

Timeline Sync

All panels are synchronized by message timestamp. Scrub the timeline once and every panel follows.

Multi-Camera

Upload recordings with multiple image topics and view all camera feeds side by side, synchronized to the same moment.

​
Fleet & Observability

Recordings viewed in the MCAP viewer integrate with Avala’s fleet management features:
  • Timeline events — Errors, anomalies, and state changes appear as colored markers on the timeline. See Events & Markers.
  • Diagnostics panel — Monitor device health metrics (CPU, memory, network) alongside sensor data. See Diagnostics Panel.
  • State machine panel — Visualize robot state transitions synchronized with the recording timeline. See State Machine Panel.
  • Custom panels — Build domain-specific visualizations with the TypeScript Panel SDK. See Custom Panels.

​
Next Steps

Supported Messages

Full reference of recognized ROS and Foxglove message types.

Panel Types

Learn about the 8 panel types and how topics are assigned to them.

Custom Schemas

Use Protobuf, JSON Schema, or ROS .msg definitions for custom message types.