How do you interact with soil sensors? - Accessories & Sensors / Soil Sensors - Rachio Community

post by rraisley on Oct 13, 2021

While I would love to have an automated system based on moisture sensors, I don’t think it is practical. Someone posted here a while ago that they got actual soil moisture charts (for a golf course or other commercial system), and had also used the service on his yard. But for most everyone, that kind of system is out of the question. I believe even getting the information was on the order of thousands of dollars per year.

What I did see in those charts, though, is what I observe in my own yard: moisture varies greatly, even in smaller areas. I have a $100 moisture meter, and get different readings several feet apart in the same zone. Yet the grass looks consistent and proper. Water can bleed from one area to another, so it /wants/ to even out, but does it poorly.

In my tests using catch cups, I found that the water applied over any zone can vary greatly (3:1 or more), and yet the yard, again, looks fine and even. The soil compensates. Some. So even if we could measure actual moisture, accurately, across a zone, the water being added is nowhere as even as we think, and would like, it to be.

Even if every zone had a 100% accurate moisture meter in it, is the meter in the right or best location? And I personally believe that people don’t know their actual Nozzle Flow within 30-50% or more. Without knowing all of that, we’d still be stuck adjusting /something/ to provide more or less water for the lawn as needed.

Anyhow, yes, I’d love to see something, but don’t think it will happen, and we’ll still have to adjust things to get them right. The moisture graph in Flex Daily may not be 100% accurate, but it describes on a per zone basis how the zone is doing, on average, or at least how Rachio thinks it is doing. I personally think that’s enough, on a properly calibrated and adjustable system.

post by drew_thayer on Oct 14, 2021

@gaustin Thanks for your input. I agree with your general principle that this kind of engineering comes down to moisture management.

When you monitor drydown, do you typically see a similar drydown curve (i.e. a similar elapsed time to some defined level of dryness) after your irrigation events when you measure various parts of your yard?

I’m curious because a potential blocker to this project could be that soil texture/composition variations in real life yards vary across such a large range that finding a characteristic drydown pattern would prove impractical. Like any modeling application we do, success with a sensor would require that we could generalize some expected drydown behavior to most people’s turf and garden zones (i.e. not move into a world with a ton of edge cases). Cheers.

post by drew_thayer on Oct 14, 2021

@rraisley Your points about inevitable variability in most yards is well taken, and is likely the largest blocker to coming up with an easy to use, consumer solution. We are going to push the hardware and software to see what’s possible.

I appreciate your anecdotes about the range of application you’ve found in your yard. This is common; I’ve found similar variation while testing on mine.

Nozzle flow is, as you’ve identified, the hardest important parameter in our system to estimate, or even measure accurately. We’ve been living in a world where we just have to assume that this rate is pretty erroneous for most customers, and I’d love to constrain that any way we can. So the idea for a sensor providing dry-down feedback is to basically get around the nozzle rate problem.

After observing our system working on a few hundred thousand yards for a few years, the applicability of a sensor arises from considering what you point out: a properly calibrated and adjustable system. You are clearly technical, detail-oriented, and patient enough to have calibrated your system well, but a large segment of customers lack the time, interest, or know-how to do this; thus we have a lot of poorly calibrated systems out there that could be optimized. Certainly some confusing elements of our UI are not helping (improvements coming!), and we’re also exploring a sensor that provides site-specific info on drydown rates as well. Appreciate the conversation!

post by gaustin on Oct 14, 2021

@drew_thayer, this dry down is highly variable and mostly dependent on the weather and ET. I was using it to match what the Rachio was doing so I could hone in my yard to your model. I still have to skip cycles and add syringe times here and there but I’ve been happy with my Rachio.

I live just north of Houston, so when the weather is hot and dry in the summer with no afternoon showers I can be very precise with my watering.

@rraisley, I’m the one you are referring to with the Visual Insight. I do believe that there can be a smaller scale application to this technology but it will definitely need to be economical for home use.

At least you are looking into the possibilities.