Inconsistent CANBUS_TIMEOUT errors occur during track following

support
1

Hi all,

I am using a modified version of the waypoint-navigation branch of the farm-ng-amiga repo. The Amiga detect objects, calculates their gps location, and creates a track to follow to get to the object. Currently, there are infrequent CANBUS_TIMEOUT errors that I have not been able to pin down a cause for. Once experienced, the robot quickly recovers, but can make for less accurate track following at times.

Resistance across the canbus is correct, and all canbus diagnostics commands I have used indicate the health of the network is fine. I suspect gRPC may be the cause of the timeout if this is getting overloaded with commands, but I don’t understand gRPC well enough to confirm this. Has anyone experienced similar?

2

Hey Patrick,

Did you try increasing the wheel odometry stale threshold on Settings > Robot Localization tab on the Brain?

We noticed occasions when the CAN bus gets busy and are testing on some settings to improve it.

3

Wheel odometry has been set to 1s for all our testing. I see the default is 0.5s. Should this be further increased?