Yes I was just suggesting , perhaps if they had longer standoff the rtk could sit higher and have better reception possibly. Mine is finally working great I have no complaint at the moment even the battery is not overcurred alarming, could it be the update has addressed it?
1 Like
3.9.29 throttles charging when hot and auto resumes if it gets too hot to charge
1 Like
I’ve suspicions on that.
It needs both antenna to have 5 L2 + RTK to compute heading.
But it only needs a good fix with 4 L2 from either antenna to know position. It does not need to prove it is within bounds - it just needs to suspect it is out of the work area by some inches and stop for evaluation. The heading, in that situation, is of no merit - and the core should not keep chugging along if an L2 with even a 4 sat fix says you’re in the middle of a street.
My fear is when they use the IMU or whatever, they ignore the antenna completely, even if one has a real fix.
2 Likes
They should absolutely use fused positioning if the RTK status is fixed and good. IMU should be trusted when not and also when GPS says it moved 20ft in less than a second (which is physically impossible).
4 Likes
@yarbo-forum @marco @bryan.wheeler @ken.w.gregory Seems we may need help with this issue again I’m thinking. From my last email R&D I’m told is looking at this again. Just now with a new core. Everything was work great, until the old core had to be sent back for a warranty issue. Ticket #114007. Thanks for any help resolving the issue.
2 Likes
Bryan, have you heard of this which happened to my new replacement core (but it also happened one time with my old core)? My Yarbo was mowing right along and when got to the end when it would normally do the boundaries I told it to go recharge as I had already done the boundaries manually. It headed for the charger and when it go a little past half way there, it turned around and headed back to where it had been when I sent it to recharge. I just watched it to see what it would do and I checked the diagnostics. The gps was good. It traveled back to where it had been, then turned around again and headed back to the charger. This time it went all the way home and began charging. Verrrrrry strange. It has not done that since.
Ken
3 Likes
That’s a new one to me. If you know the time/date I’d open a ticket and post the number here for them to investigate it.
2 Likes
So in other news,
It almost looks like the initial perimeter lap uses a different codebase from the 2nd lap, or the interior process.
Anyone have any advice for dragging a 200lb piece of crap out of an 8 foot ditch with only one hand while working a phone with the other?
Literally ZERO reason for it to be off course when it started that turn, way too early, let alone for it to not call out of bounds. The entire lane that it CUT was out of bounds, one if not both GPS units knew it, and it should have faulted the job immediately.
1 Like
Wow, that’s not good. Seems you got through it ok though? Were you watching it in the app at the time it did that? Wonder if it was a multipath error and then slipped down the hill when it was turning?
2 Likes
I hear Yarbo has a tow mode. Maybe that will work for you. 
4 Likes
I had the map open in BlueStacks, and noticed it was “navigating an obstacle”. Looked out a window and couldn’t find the core, anywhere, lol.
I wish it was a multi path error. That section of that yard is clean and perfect for reception.
The corner where it started over that boundary, however, is overhung with conifers - but this fuxup entails a full 40 feet of clear sky with a working-speed preference on SLOW. So, I think a couple of things are going on with this failure.
- Navigation for the initial border lap is different than the rest of the job, and it contains bugs that are not present in the other code
- When the core uses PPVS, it completely ignores a good GPS lock for true position, and ignores bread-crumbs for true (effective) heading.
- If one of the M1’s casters are out-of-bounds in unmanaged weeds, I’d wonder if the yaw is subtle enough that PPVS / IMU either discards it or does not realize it. Q.v. this popular off-road hangout for Yarbo, over and over:
If you zoom in a bit, you’ll see the core. Notice it even goes around the tree to get there. How nice.
At any rate, it’s somewhat similar to the photo in the prior post, as if it gets pulled out of the lane and doesn’t recognize it on the initial perimeter lap. The assumed heading was wrong, and the assumed position was off by a LOT. In the image below you can see where the turn should have started, vs where it did begin.
4 Likes
He’s off in the weeds…. sort of speak
2 Likes
I’m sure it goes without saying but I’ll say it anyways. I hope you have a ticket so they can pull the logs and fix this. Quite puzzling. I assume it’s been doing this area fine up until this point? Certainly if it knew it was out of bounds then this should NOT ever happen. Like you said, throw an error!
1 Like
Nah, this is semi-recurring behavior with those two specific area perimeter laps. They are why I asked for a “no perimeter” option back… back in… uh, a long, long time ago.
I find there’s no point in making tickets for firmware behavior. The build they’re working on has little to do with the one we have in production. I might as well submit some bug reports for Windows 3.1, if that makes any sense.
Even if it was relevant, there’s still no reason. Production would not see a firmware that addresses this until Q3 2026. Or maybe Q1 2029. Even Oracle has a faster release cycle; they run separate bug fix timelines vs feature timelines, at least.
So, “no point.”
6 Likes
Ok. That was my favorite comment of the day!
Sorry about your perimeter problem though. I have not experienced the same issue with my unit.
4 Likes
I didn’t know we weren’t supposed to be submitting bug reports for Windows 3.1 any longer. Explains why it’s not getting fixed maybe? 
3 Likes
I am inclined to let sleeping dogs lie. No problems since.
2 Likes
What are your PPVS settings set to? I’ve been messing with mine and things start getting hairy much above 70 feet on the border. I’ve been able to leave the in area settings maxed. I do think if a track slips it gets off track a lot easier, excessive turning or back and forth dances add up quick.
3 Likes
Until I gain confidence in PPVS, I’m keeping my settings very conservative. I’m not interested in Yarbo accidentally finding its way into the street or damaging any of our vehicles or property.
So far, I haven’t had any stray Yarbo incidents to speak of. Knock on wood.
4 Likes
I marked all my dangerous areas no go zones, “just to be safe.” That way it doesn’t try to turn around outside them at all. The pattern there though makes it look like either it slipped or decided that it was an acceptable place to go.
2 Likes