Ros2_medkit + VDA 5050: bridging SOVD diagnostics with fleet management
New plugin lets ROS 2 robots report detailed faults to fleet managers using the VDA 5050 standard.
Selfpatch AI has unveiled a significant integration for its ros2_medkit diagnostic framework, creating a bridge between detailed robot health data and the VDA 5050 standard used for industrial fleet management. The core innovation is a new Service Interface plugin that exposes medkit's structured diagnostic data—including its entity tree, faults, and capabilities—via standard ROS 2 services like `ListEntities` and `GetEntityFaults`. This architectural shift means any ROS 2 node can now programmatically query this diagnostic information, moving beyond the traditional SOVD/REST client model.
In a practical demonstration, the team built a separate VDA 5050 agent process that consumes these new services. This agent handles standard MQTT communication with a fleet manager for orders and navigation via Nav2, but critically, it can now query ros2_medkit to enrich the minimal VDA 5050 error reports. When a fault like a LiDAR failure occurs, the fleet manager receives the basic error code for operational decisions, while engineers get a full SOVD snapshot in medkit's web UI, complete with freeze frames and rosbag data for deep debugging. The plugin's design keeps the diagnostic system agnostic of the fleet protocol, making it reusable for other applications like behavior trees or visualization tools.
- New ROS 2 service plugin exposes ros2_medkit's full diagnostic tree (entities, faults, capabilities) for programmatic access.
- Separate VDA 5050 agent translates rich SOVD fault data into the standard's minimal error reports without breaking either protocol.
- Demo on a ROSMASTER M3 Pro showed a LiDAR fault triggering both a fleet manager alert and a full diagnostic snapshot for engineers.
Why It Matters
This enables industrial robot fleets to maintain operational efficiency while giving engineers the deep diagnostic data needed for rapid repair and maintenance.