[Feature request]: Figure out what to do in the event of loss of odometry from a single agent.

Before proceeding, is there an existing issue or discussion for this?

• I have done a search for similar issues and discussions.

Description

A critical challenge in production environments is handling robots that unexpectedly go offline or lose localization. The current prototype implementation handles connectivity loss by stalling the fleet, which is unsustainable for real-world deployments where intermittent network dropouts are common. A robust framework is required to manage these scenarios effectively.

Implementation Considerations

The optimal behavior should be highly configurable, as different operational use cases demand distinct solutions. For example, in a high-density, highly controlled environment with a robust wireless network, it may be appropriate to issue a localized stop-work order and reroute other robots so a technician can safely retrieve the non-responsive agent. Conversely, in public service deployments like hotels or hospitals with minimal human support, controlling Wi-Fi quality is more challenging, and robots may continue operating normally despite intermittent connectivity.

To accommodate these varied scenarios, we must design an API that empowers system integrators to define and configure the appropriate error-handling behavior for their specific deployment needs.

Alternatives

No response

Additional information

No response

The issue names no files, tests, or entry points, so first map the prototype's connectivity and localization-loss handling before proposing an API. Done means a configurable framework lets integrators choose appropriate behavior when one agent loses odometry, with tests covering the supported handling scenarios.