Improving insight into the robots behaviour.

[SeaTechRC]

Hello,

In order for a user to see why the robot has stopped the automation,
the reason should be kept and shown under the start/pause and stop button.
Otherwise, if the robot stops while being unattended, the user will not know what caused it.

Same would go for temporary stops (RTK fix lost, bumper pressed, etc.).

It could go even as far as showing the current state of the implement and the navigation, since both are state-based (or theoretically could be).
This could also be very nicely visualized with mermaid diagrams.

[rodja]

Yes, sounds great! We already have such a state representation and logging view for other robots.

[rodja]

The state-representation can be implemented with rosys.analysis.track. Additionaly feeding a ui.log from a history of notification events could be helpful:

class LogMonitor(rosys.persistence.PersistentModule):
    MAX_LINES = 100

    def __init__(self) -> None:
        super().__init__()
        self.lines: deque[str] = deque([], self.MAX_LINES)

        self.NEW_LINE = rosys.event.Event()
        """a new line was added to the log (argument: line)"""

        def handle_notification(message: str) -> None:
            line = f'{datetime.now():%m/%d/%Y %H:%M:%S} {message}'
            self.lines.append(line)
            self.NEW_LINE.emit(line)
            self.request_backup()
        rosys.NEW_NOTIFICATION.register(handle_notification)

    def backup(self) -> dict:
        return {'logs': list(self.lines)}

    def restore(self, data: dict) -> None:
        self.lines = deque(data.get('logs', []), self.MAX_LINES)