Hi there,
I’ve noticed that yarbo’s routing isnt’t efficient, meaning:
In linear mowing mode, when one “subarea” is completed, yarbo does not start the next nearest area (green arrow) but moves to the furthest away “subarea” (red arrow). It would make more sence to use the nearest area to save battery and mowing time.
Also notice, that battery is good and the docking station is on the right side of the map, so there is no need to mow near the docking station:
Common complaint and request. I believe this is on the roadmap if I recall correctly from some of the FB Yarbo responses. Not sure how soon though.
Thank you for bringing this to our attention.
As Bryan mentioned, route planning optimization is something we’ve heard feedback on quite frequently, and we completely understand your need for a more efficient mowing path. The behavior you described—where Yarbo moves to a farther subarea instead of the nearest one—is definitely something that can impact battery use and overall efficiency.
We’ll share your feedback with our product team so they can take it into consideration for future improvements to the route planning system. Your input is truly appreciated and helps us make Yarbo better for everyone.
If you have any more suggestions or questions, please don’t hesitate to reach out!
Same here. Don’t like that at all. Sometimes it travels hundreds of meters because of unmowed area. Few examples.
Check this out…craziness
Mower goes back and forth all over. In the circle note that the path is 90 degrees out and least efficient to boot. This is on parallel path.
Yes, admittedly they have stated that path planning is still a work in progress and has a lot of room for efficiencies and optimizations. Hopefully one day soon we will see an algorithm that would mow closer to the way a human would approach these and go around the obstacles in a more efficient manner as well as plan a more appealing look and efficient route overall.
There is definitely room for improvement.
I think it’s odd that the mower logic is pretty bad when the snowblower logic seems much more efficient. With a snowblower you can’t just randomly drive around an area before it is blown…seems like the same logic could work with the mower more effectively than what it uses now. That being said, we had a mild winter so I didn’t get a chance to use my blower last year so the logic could now be worse than it was last time I ran the snowblower.
They’ve been working on the snow blower since 2015 and the mower since mid 2023. It’s not hard to see why there is still some finer details to flesh out for the mower.
Efficiency would be ideal but I just want the damaging behaviora addressed. Can anyone confirm the following zero point turn?
It seems to defaults to a zero point turn on the first boundary perimeter lap.
No mater how you create paths to approach the charger, it makes unnecessary zero turns to dock.
When moving between perimeters of no go zones it straight lines to the same point each time and depending on approach direction, zero turns to start the no go zone boundary.
Yes, this is what it is currently programmed to do. I know it’s high on the priority list to fix.
It is very hard to me to find a pattern. Even when smart turn is clearly selected, when it does the diagonal, very often it does at least partly a zero turn, even if I am 100% there is enough room for a smart turn. I don’t care if it takes the robot 10 movements to turn, we should have the option to forbid completely zero turns… It is not that bad, but my perimeter is not as green as the rest because of that…
And yes when going to the dock, it does zero turn all the way… I moved the dock on asphalt because of that… It completely destroyed the lawn in front of the fist location I installed the dock… in under 48h… It is still recovering from that experience 1 month later…
So to me to make it work, I created as many paths as possible between areas on “non-grass” areas such as entryway, concrete pathways, etc, enlarging the border of some areas so the pathways ends at that border that is still on concrete or asphalt so the first non authorized zero-turn is done on something hard…
I still have some areas where this can’t be avoided, so I make sure the pathways start and end in less visible areas such as behind a garage, hidden by a tree or something like that… But sure thing, your lawn will be destroyed at the start and end of pathways… And if your pathway has a sharp turn, it will also destroy the lawn there.
But if you are careful when you create the pathway to avoid sharp turns, make the pathway on hard ground or less fragile ground, and start and finish in hidden areas or hard concrete/asphalt/dirt ground, then what is visible is superb…
While the Yarbo mower was still in development Mammotion was dealing with every single one of these software issues. Either no one from Yarbo was watching, or they pushed it out the door anyway.
Mammotion also doesn’t have to deal with snow blowers, leaf blowers, smart assist modules, trimmers, 2 mower modules, a data center, oh and keeping all of the 2023 M1/B1/S1 HaLow base, LoRA base, and universal body also relevant and working too. I don’t envy their task. Doing pretty well all things considered.
I get it, they bit off a lot, but don’t claim features that aren’t real. The marketing materials are STILL doing it! The mower issues we’re all dealing with were all well documented and could have been avoided. The snowblower is great, but none of the other modules do what was claimed.
Not trying to make excuses for them because the marketing should absolutely be realistic. Asterisks are sufficient with rough estimates on timeframes for release. They have been working on this. But, at the same time, it’s a very competitive market and they are a small company trying to make their mark.
I learned very quickly trying to get information about things during the preorder campaign that the marketing was a lot of hype. But, the live streams, user forums, pre-sales, and after sales support team were all very straight forward with what it could currently do, what it would do at launch, and what it WILL eventually do. It’s been a journey. Honestly, I’ve had a lot of fun on it with them.
Does it mow my lawn, yes it does. Does it do it well? Not always the first time. But, that second run gets it. Is it getting better? Almost every release. Am I happy that the trimmer is rear mounted and not mounted on the front like originally advertised? Yes and no.
I bought thinking the M1 was going to run the edger and trimmer with the blower mounted on the rear. This is no longer a reality. But, I think the redesign of the trimmer is the most practical and logical choice. It is a better design. But, it’s not what I was sold for sure. I realize this was also just a concept at the time of the preorder. I’m hopeful that maybe in the future a different mounting method or bracket may make it possible to have the trimmer and blower mounted simultaneously like I was originally counting on. No one wants to swap modules just to blow off the driveway after mowing and trimming.
I also wasn’t happy the leaf blower didn’t articulate left, center, and right like the proof of concept showed. But, I think the redesign makes sense and the detachable hose was a nice bonus. It’s also way more powerful than originally advertised. So, as an early adopter of these conceptual products (at the time of preorder the 2024 was still being engineered). I think they made some tough decisions and in the end released the best product they could in the timeframe that was allotted (which was even longer than they wanted or hoped for).
I get challenges of a more dynamic system and the competitive market, but updates should bring improvement NOT damage.
Yarbo was cutting about 95% of what I had expected it handle as I gave up weak GPS spots I. Favor of consistent completion.
Now I have had to remap with smooth curves and give up nooks due to the introduce behavior of the last update.
It is covering less yard now ~80% due to the damage it was causing. This is the wrong direction!
Agreed. I think they felt it was necessary to prevent accidental damage while they work on improving it. Good things are coming soon!
You are right with the fact they have a lot of devices variations to maintain, but it supports our point. They might wish to hold a bit on the new products release and nail what they already released…
Again, I am satisfied and I would buy again, but I have been a lawn robot owner for 8 years now and there are things the older generations of robot already nailed a while back, both at the software level (extremely almost not useable scheduling, massive bugs in the app) and “hardware” (I put in there poor navigation, pathways, etc).
Now, the RTK, fusion navigation with computer vision and so on, this is something we can be more lenient as this is breaking edge techno… But optimized routing is a SOLVED problem. It is called “graph optimization”. There are well published algos to do that (A*, Bellman-Ford, Dijkstra…). To apply that there are widely available libraries (including in C++) to do what is called SLAM (Simultaneous Localization and Mapping) and it is NOT computationally intensive. It is either they are really really badly using those libraries, or they don’t at all apply any optimization to the routing. This is probably what irritates me the most… I understand having bugs in an app… I can even forgive screwing up localization settings (the dreaded foreign language special characters bug that literally kills the app). This is pure oversight and it happens all the time… But selling an “intelligent” GPS-based device and not applying basic graph optimization algos is mind blowing… (or applying it so poorly that we don’t even notice it is there).
And regarding the grass weapon of mass destruction? The robot is doing great when not doing those d**n zero-turns! They just have to stop that. NOW. Like in “realease a hotfix yesterday to stop zero-turns when we select smart turn NOW” sort of now… Not in in 2 months… Now… @Yarbo-Forum
