4.9 eCat update, ‘Mission Controller’ now supports Redis communication. You can remotely send messages to the Redis server, and once VC receives the message, it will immediately dispatch tasks to AMR, check RedisScript for more details.
If possible, can you upload a video? Unable to access YouTube, thank you very much
Sure
1 Like
I have a new idea, can tasks be sent remotely through the HTTP protocol? Because many AMR hardware upper RCS systems now communicate through HTTP protocol, if feasible, it means that in the future, VC can perform virtual debugging with RCS from many AMR manufacturers
Can you share your VC file?
@idkfa If the scene consists of multiple AGVs, how can different tasks be assigned through the Mission Controller at different times?
I have currently tested and found that no matter how many AGVs there are in the scene, if the AGV fails to complete the task and the Mission Controller issues a new task, the AGV will report an error. The new task can only be issued after all AGVs in the scene have completed the task
You need to create different missions to execute collect and devlier tasks at the same time
{
"missions": [
{
"id": "Mission #1",
"enabled": true,
"priority": 10,
"activator": "Repeated",
"delay_between": 0.0,
"start_time": 0.0,
"limit_occurrences": true,
"occurrences": 100,
"accepted_resources": "Any",
"steps": [
{
"type": "Collect",
"scope": "Any process by id",
"process_group_id": "Feeder",
"optional_step": false,
"collect_all": true
},
{
"type": "Deliver",
"scope": "Any process by id",
"process_group_id": "Sink",
"deliver_all": false
}
]
},
{
"id": "Mission #2",
"enabled": true,
"priority": 10,
"activator": "Repeated",
"delay_between": 0.0,
"start_time": 0.0,
"limit_occurrences": true,
"occurrences": 100,
"accepted_resources": "Any",
"steps": [
{
"type": "Collect",
"scope": "Any process by id",
"process_group_id": "Feeder",
"optional_step": false,
"collect_all": true
},
{
"type": "Deliver",
"scope": "Any process by id",
"process_group_id": "Sink",
"deliver_all": false
}
]
}
]
}
@idkfa Simply put, for example, when an AGV is executing a task, the task controller combines different triggering conditions to form different new tasks and sends them to the mobile robot controller. However, there are many possible combinations of these contents, which are unknown in advance. If there are 100 possible tasks, manually creating so many tasks would take a lot of time, and later adjustments would be needed. This project would be huge. I hope the task controller can issue new content tasks at any time without interrupting the AGV in the task. Can you understand??
I didn’t read your script, but I think you should save your new missions somewhere else instead of in Note Behavior directly. Update Note Behavior only when the current missions are finished
@idkfa It doesn’t matter where the task is stored, just like the video case you posted:
Assuming there are two AGVs in your video case, when the simulation starts running, the first task (moving from A to D and then to A) is assigned to AGV 1; Before AGV 1 completes its task, it suddenly needs to issue a second task (moving from B to C). Theoretically, AGV 2 should be assigned to execute the second task, and both AGVs 1 and 2 can complete these two tasks independently without affecting each other. However, after testing, I found that when the second task is issued, AGV 1 will report an error and stop running, while AGV 2 will also run the second task
I don’t know if you can understand my expression, or if you can follow the test above and you will know
@idkfa Because in many scenarios, the task steps of AGV are variable and the triggering time is uncertain, it is impossible to predict in advance and manually create all task steps in the Mission Controller component. Different task steps need to be created in real-time based on conditional logic.
The mission controller was not originally designed to dynamically change missions during simulation. That’s why I said you have to wait until the mission is fully executed before updating Missions::Definitions. However, if you update Missions::Definitions before the mission is complete, it will cause errors. If you want to do it this way, you might need to redesign different components.