Snow blower slope improvement

I’m using my yarbo snowblower with the slope pattern but I see some room for improvements.

When going back up the slope the snowblower is being tilted down (the same setting as while working?) but it’s not snowblowing. It frequently keeps getting stuck doing this. I don’t know why you can’t select that it should blow snow also on the way up? There should be an option for that. For the time being, at least it should lift the snowblower module higher up so it doesn’t keep getting stuck.
But I think the possibly best solution would be if it backs up the whole slope? In my case the slope ends at the public road and I’m reluctant to let the yarbo go close to the road since it does these tank turns. Doing tank turns in the piles left from the snow plow has the yarbo frequently getting stuck. So if I could choose I would want to Yarbo to work it’s way down the hill until it reaches the public road. There it lifts the snowblower and backs up to the top of the slope. Or it backs up a few meters, into a previously cleared area, before it turns there (not in any uncleared area) and goes up the slope with the snowblower lifted high, or running the snowblower to clear any leftover snow from the previous pass.

Side note: I have a gravel driveway and keep some cm of hard packed snow as working surface for Yarbo throughout the winter.

4 Likes

Thanks for sharing your detailed suggestions and real-world experience with the slope pattern. Your feedback is very helpful, especially regarding the return behavior on slopes and the challenges around turning and getting stuck. I’ll make sure to pass your input along to our product team for their review and consideration. We really appreciate you taking the time to share these ideas.

4 Likes

I have made some further experiments with the slope mode and some heavy snow here. Let me add todays insight, and a video https://youtube.com/shorts/VQXRnR4vgus?feature=share

First some background information:
My slope is a gravel driveway of slightly varying with. There has to be a thin, but hard snow layer to cover the gravel and occasional grass. Also it’s not possible to design a square shaped area for snow removal. Think of the area as banana of varying thickness, where one end is the top, while the other end is the bottom of the slope.

For todays testing I moved some wind compressed snow from the last storm by shovel into the previously cleared driveway. It had accumulated in an area with even more uneven surfaces, that can’t be cleared by Yarbo. I set the working speed to the slowest possible, and the intake effect all the way to Performance (it says 30).

Things that need to be addressed:

  1. Pattern:
    a) Because the sides of the slope are not parallel, straight lines, in the direction of the slope Yarbo makes a pattern where it needs to make turns in the middle of the slope. It’s very inconvenient, since the Yarbo tends to get stuck when trying to make turns in a relatively steep slope with heavy snow. After all Yarbo suffers of a complete lack of self awareness, and doesn’t know that it can’t turn its rear or the snowblower into areas that haven’t been cleared yet. So I suggest that all the lines of the pattern run from top to bottom of the slope. Either with a varying overlap or by distributing the overlap evenly between the lines. Think of an hourglass, stretching from top to bottom of the slope. All lines will touch in the middle of the hourglass.
    b) For some reason Yarbo doesn’t start at one side of the slope and gradually work it’s way towards the other. After doing 3 lines on one side it’s suddenly deciding now would be a good time to clear a line in the middle and work from there. Only to later return to the side and throw snow onto the middle area it has previously cleared. No one ever clears an area like that with a snowblower. Start on one side and work from there all the way over to the other side of the slope. It’s hard enough to push that first line, where it has to take on the full with of the snowblower, once down the slope. And jumping from side to middle, back to side and so on just looks stupid.
  2. Snow blower encountering too heavy snow
    When going down the slope at minimum speed and maximum performance and encountering too much snow Yarbo responds to beginning to get stuck by accelerating it’s drive belts FORWARD, just to make sure it’s absolutely sure to spin itself down into the snow and get stuck. What it should do instead is lift the blower to the maximum angle, and keep at the lowest possible speed. If that doesn’t help it needs to stop right away, revers back up the hill to before it started to struggle, lower the blower to the desired setting and try again at slow speed.
  3. Going back up the slope
    I have already addressed the problem of returning up the slope with the none operating snow blower being pushed trough the snow. It’s really a big issue for me, and there should at least be an option to enable snow blowing on the way up. Yarbo keeps getting stuck on the way up. And when it gets stuck it starts the blower for a few seconds. Now if it has started to dig itself down it has a somewhat useful solution to that, except it stops spinning the snowblower just when it’s about the pass the problematic area and reverses back so it’s stuck in about the same place as before. By running the snowblower constantly on the way up, this wouldn’t be an issue. The way it’s now it’s quite frustrating to watch.
  4. Making turns
    I need to repeat what I mentioned above. Yarbo makes turns into areas where it hasn’t removed the snow. That might work for people who get 5-10 cm of snow. But here, where we can have 0,5-1m within hours it’s impossible to turn that way. Yarbo must develop an understanding of its circumference and clear the whole area it needs for a turn before turning there. The way this would be done is by reversing in it’s own track and going forward again in a fan-shape until it has cleared an area that’s wide enough to make a donut turn, if that’s what it needs. Only then it’s okay to make a turn without getting stuck. By attempting to turn the way it does not only does it get stuck, but it also leaves a lot of residue snow where it turns. Sometimes it also ends up outside de designated working area because it starts slipping sideways in the slope.

So to answer my question in the video, I think Yarbo would have been able to remove all of that snow if it was better programmed. But the way it’s working now it had no chance. I had to pull it unstuck several times, and ended up with a somewhat wavy driveway because Yarbo compacted snow that it should have removed very unevenly.

Take not of my surroundings in the video. I have a lot of experience with heavy snow removal and I want Yarbo to work. But for that it has to become much smarter. I’m happy to have a whatsapp call and discuss my suggestions with whoever is programming this. So the slope function starts working in a useful way. I’m already getting tired of going outside to get Yarbo unstuck.

4 Likes

Made another two obersvations today:

When Yarbo is done pushing snow up the hill it will turn on the snowblower and blast whatever it has gathered in some direction. In my case it’s not in a direction I’ve told it to blow snow. The core then starts to turn, distributing snow in all directions before it has aligned with the new work direction.

And all the pushing around of snow uphill is starting to develop waves like the sand at the beach. Look at the shadows in the snow between myself an the Yarbo. Some more passes and Yarbo started to get stuck in those waves, despite of this being the most flat part of the slope.

2 Likes

I have a couple waves forming like that where the Yarbo transitions from the side yard, over the front yard to the driveway. What height settings are you using? I started the season at ~.8” but lowered it to ~.5” now that there’s a base…was thinking about going even lower.

1,1 cm

Thanks for sharing your latest observations and detailed feedback. I’ve shared your feedback with our team for consideration. We truly appreciate your input as we work to improve Yarbo’s performance!

1 Like

100% agree that the blower should lift all the way up when going back up the hill. if not, it gets stuck on the…slope. and then raises, goes forward, stops, goes back, lowers, gets stuck on the slope….rinse and repeat….

3 Likes

Yeah it’s somewhat infuriating when you get a track slippage alert, you open smart vision/controller, the height is 1.6” or lower so you raise it up and just backup 3’ and hit resume and all is fine in the world until the next weird zero turn​:melting_face::melting_face::melting_face:

2 Likes

Greetings,

I had my first reasonable snow fall last weekend. I give the results a C grade, which is honestly promising for a first run. Overall I was very happy and believe there will be good OTA solutions or workplan workarounds available to improve the score.

I’ve replied to this post as most of my observations are related to slope mode, and have already been discussed in similar threads. Consider my post a +1 to the need for slope mode improvements for the devs.


-details–

rating factors (10 scale):

  1. Setup: 8
    1. Definition: Relating to the amount of mapping and workplan logic implemented by the user to achieve the desired result.
    2. Notes: The app provides a reasonable interface and is generally navigable. I made 3 areas, 3 paths, and settings for each area.
  2. Execution: 6
    1. Definition: Did it do the job as intended and expected?
    2. Notes: Yarbo partially executed the workplan as designed. The results were less than ideal given large piles of snow left at boundaries, large piles left at overlapping sections of mapped area, the need to improve turning with deeper snow, and challenges with editing a map vs. remapping - some old ghost sidewalks (? it shows I removed them - and I did a battery cable pull for 5 min) were conflicting with the slope lanes, resulting in yarbo following lanes with perpendicular ghost sidewalks and then navigating using the sidewalk to change lanes instead of continuing past the ghost sidewalk and restart the top, while marking completeness (incorrectly) on the plan view in the app.
  3. Automation: 7
    1. Definition: How much manual intervention or manual snow clearing within the area was required?
    2. Notes: The majority of the snow was removed! But I had to shovel the piles mentioned previously and the section that was missed completely because of the ghost sidewalk. Also I needed to re-install the track studs due to slippage, and I helped it get unstuck once all contributed to a need for ‘hands-on’. I am manually setting the starting point as I don’t have a static spot for the charging base yet. This score will improve quickly with OTA improvements and workarounds to the mapping and workplan - and a permanent home for the charging base.
  4. Intuition: 8
    1. Definition: Was it intuitive to use the app and hardware, and to troubleshoot challenges or mistakes?
    2. Notes: I won’t go into deep detail here; I was able to achieve the results minimal need to look up directions, or learning more about configuring settings. Much of the score was influenced by initial research on the forums to create a strong start, and learn what symptoms stem from which causes.
  5. Results/Style: 6
    1. Definition: Were the aesthetics of the execution making me confident of my investment? Or, was yarbo riding the struggle bus to the point neighbors were feeling sorry for it? Did the driveway prove functional though embarrassing, or are neighbors in amazement and abuzz?
    2. Notes: Visually speaking, the result is rougher compared to my neighbors manually operating their snowblowers. The novelty from the neighbor perspective is high until they see the initial results, and then I need to sell copium that ‘I’m in the initial testing stages’ or ‘I’m trying some new settings to improve it’. I love tech, and this is awesome. While I did intervene and spend a lot of time on initial setup, the time not spent clearing the driveway was well worth it. I expect this score to improve greatly with OTA updates and my manual workarounds for workplans.

Echo of feature requests that I can confirm would be useful based on observation:

  1. Top side full length clearing before starting slope runs, 1 to 3 yarbo width lanes
    1. Option: auger height + second pass. This will help turning and minimize snow shrapnel being strewn about.
  2. On return up the slope, use a cleared lane with margin (+.5 to +1 lane)
    1. mine kept tracking (left towards snow throw direction) into an uncleared lane, turning Yarbo into a snowplow going uphill and leaving a large line of snow piles on the top that don’t get cleared.
  3. On return up the slope, lift the module (option: height) + operate (quiet mode) (traversing previously cleared area in #2)
  4. When stuck going down the slope, backup slightly (TBD) and retry 3 times each with progressively increasing auger height.
    1. On progress success, return auger height to plan.
  5. A ‘cleanup perimeter pass’ would be great at the end of clearing alll slope lanes, as part of pathing.
    1. This can be achieved in a workaround through sidewalks, but I noticed issues with pathing when sidewalks overlapped the area even though they weren’t in the workplan.

My last thought is there are many similarities between the path generation and 3D printing slicer software. I think Yarbo could make major improvements to path creation by emulating patterns from Klipper (e.g. inside borders-for turning, outside borders-cleanup pass, infill patterns, speed, flow rate, etc). I think the best result would be if Yarbo would create the ‘Yarbo Slicer’, and allows users a higher level of control over the path planning to account for our unique and specific area characteristics (e.g. mapped auger height profiling vs. one auger height).

2 Likes

Hi there, thank you so much for sharing such detailed feedback on your first run and for taking the time to document your observations so thoroughly. We really appreciate your thoughtful suggestions regarding slope patterns.

I’ll make sure to pass all of your feedback to our product team for consideration. In the meantime, we’ve already made some optimizations in the upcoming release based on input from other users with similar use cases. We look forward to hearing your feedback after the update and continuing to improve the experience together.

4 Likes