Task execution with intermittent robot connectivity

To answer this, I’ll search the web for “site:discourse.openrobotics.org/c/open-rmf/open-rmf-general/103 Open-RMF intermittent connectivity robot task offline charging station”.

To answer this, I’ll search the web for “site:discourse.openrobotics.org/t Open-RMF send task robot disconnected execute offline reconnect charging base”.

Ciao, ho verificato: non emerge un thread esattamente identico al tuo caso d’uso, cioè robot con connettività disponibile solo alla base di ricarica, task inviato quando è online e poi da eseguire anche offline.

Ho trovato però discussioni correlate su:

• gestione di robot offline / perdita connessione: Robot lost connection
• uso di override_status("offline") e decommissioning: Interrupting RMF control / robot offline status
• aggiunta/rimozione graceful di robot tramite commissioning: Gracefully add and remove robots
• charging station per robot: Defining a charging station for each robot

Questi thread sembrano trattare soprattutto cosa fare quando un robot perde la connessione o come evitare nuove assegnazioni, ma non chiariscono pienamente se RMF supporti un modello in cui il robot riceve un task mentre è connesso, poi lo esegue offline e sincronizza lo stato al ritorno in base.

Puoi incollare questo testo nel forum:

Question about task execution with intermittent robot connectivity

Hello Open-RMF community,

I would like to ask for guidance about a use case involving robots with intermittent network connectivity.

In our scenario, the assumption is that a robot only has network connectivity when it is at its charging station / home base. While the robot is connected, we would like to send it a task through Open-RMF. After receiving the task, the robot may leave the charging station and lose network connectivity, but it should still be able to execute the assigned task autonomously. Once it returns to the charging station and reconnects to the network, it would then synchronize its status / task result back with RMF.

So the intended flow would be:

1. Robot is at the charging station and connected to the network.
2. RMF assigns or sends a task to the robot.
3. The robot acknowledges and stores the task locally.
4. The robot leaves the charging station and loses connectivity.
5. The robot executes the task while offline.
6. The robot returns to the charging station.
7. Connectivity is restored and the robot reports the final task status back to RMF.

I searched the forum and found some related discussions about offline robots, lost connections, commissioning/decommissioning, and charging stations, but I could not find a discussion that exactly matches this operating model.

My questions are:

• Is this kind of offline task execution after task dispatch supported or compatible with Open-RMF’s task and fleet adapter model?
• If the robot disconnects after accepting a task, how does RMF expect the fleet adapter to behave while the robot is unreachable?
• Would RMF keep the task active while no state updates are received, or would this eventually cause task failure / traffic schedule issues?
• Is the recommended approach to implement local task persistence and execution inside the robot/fleet adapter layer, and then resynchronize with RMF once the robot reconnects?
• Are there any best practices for this type of deployment, where connectivity is only available at the charging station?

Any advice on the recommended architecture, limitations, or examples of similar deployments would be very helpful.

Thank you in advance.

1 post - 1 participant

Read full topic