I looked for a related topic but didn’t find anything too close. So, here goes.
I am still working on planting issues, But I thought I had the watering sequence working. However,
the watering seems to be erratic sometimes. The main issue is that it is turning on the water then moving on to another location without turning the watering off. If I stop the sequence, and check the group sequencing pattern all looks in the correct order for (“X/Y Ascending Order”), and then restart the watering sequence everything is working as expected. Not sure what gets it into its funky watering mode but will run it more times (haven’t planted anything yet). I have two Farmbots and have seen this behavior on both bots; so assuming it is a software issue.
@dlmcshan Can I see a screenshot of your sequences? Are you calling CONTROL PERIPHERAL to turn OFF the water in the same sequence that moves to the plant’s location?
When you execute a sequence against a group of plants (or assign a group of plants to a Farm Event), FarmBot OS will essentially “expand” the operation into multiple calls.
So if you do something like this:
Or something like this:
FarmBot OS will actually duplicate the sequence run N times (where N is the number of plants in the group). Because of this, you must ensure that CONTROL PERIPHERAL is turning the hose off within the watering sequence- do you think this could be the issue?
A screenshot would help me identify the issue. Please let me know if this helps.
Here is the sequence I was using.
which expands to:
and the Map for all plants is:
It was going from 10 back to 8 with the water on…when misbehaving. A second try after checking the group sequencing delivered single shots to all the plants.
I had a thought on this issue overnight. What if it is just an operator error in that possibly I started the same sequence twice. I know that is an issue turning on/off peripherals manually. Press once nothing seems to happen because of the delay; so you press again. It usually takes several iteration to synchronize.
So if it were processing two overlapping sequences, would the two sequences commands get interspersed on the FarmBot? That would explain the motion during the watering on and jumping back to a previous location.
I can imagine that this is allowed if you are running independent sequences for measurements like temperature or squirrel detection (which BTW aren’t much bothered by gantry motion!).
Thanks for sharing. This helps me understand the problem. I don’t see anything that immediately stands out as user error- it looks like the sort of sequence I have run on my Genesis many times without issue. Although it could still be a configuration/usage issue, it also might be an unreported problem with the software itself.
This has been a longstanding problem (I am spending time this week to re-write the firmware handler in order to fix this and several other firmware-related bugs). Still, generally speaking, even if a sequence is run twice, the sequence steps should not get “jumbled.”
Here are some next steps we can try:
- Reboot the device by unplugging it from the wall and plugging it back in (rebooting from the web app might not clear out transient firmware issues).
- Give the device enough time to boot up to avoid accidentally sending the firmware a message before it is ready. Normally this shouldn’t matter, but it might not hurt to be extra cautious until we find a root cause.
- Push the “RUN” button on the sequence exactly once. If the sequence does not run, wait at least 20 seconds before pressing again. Does the problem persist?
Although we haven’t seen any bug reports of this nature, it is still worth investigating whether the problem is in the FarmBot software itself. The next option would be to:
- Turn off OS Auto Update in the device settings page.
- Download an older version of FarmBot OS: Version 13.0.1 for FarmBot Express
- Perform a “hard reflash” of the device using Etcher.
- Attempt to rerun the sequence. Does the problem persist? If not, you may have found a bug in the latest version of FarmBot OS. Please let us know if this is the case.
Later today, I will remotely access your device to see if I can find any other hard-to-spot issues.
Please leave your device online during this time. Otherwise, I will not be able to access the device.
Hello again,
I did one more watering sequence test. I started and all plants watering sequence. It was working properly but about half way through, I started a second sequence. So, that definitely is the problem with it moving to the start location and back to where it was and then back to the start turning on the water and then heading back to the first sequence location with the water on.
So, now that I know what the issue is I will just try to be careful when I manually start a sequence to not overlap a running sequence. I assume this is a non problem when the sequences are scheduled in a regimen.
I will try this with an earlier version of the firmware tomorrow and retry the same experiment.
Thanks for your help.
I’m glad you got it figured out @dlmcshan. This is still unexpected behavior that needs to be fixed on our end. Under normal circumstances, the jobs should be queued on a first-in-first-out basis.
I will make a note of this for a fix in a future release of FarmBot OS. Thanks for reporting and please do let us know if you need more help or find new information related to this bug.