Moisture Sensor Integration


#1

There’s a bunch of threads on this, but none of quite solve the issue of adding cheap, accurate, and commercially-available soil moisture sensing to Rachio.

This is the ideal way to “close the loop” of the Rachio soil-moisture guesstimation.

Edyn seems like the only available option (via IFTTT), but it’s relatively expensive and the reviews suck (e.g. it reports the same moisture value no matter what). And it doesn’t actually use the moisture value to feedback into the Rachio algorithm.

My plan is to use the $14 Xiaomi plant sensor (sometimes called “Flora” or “Miflora”) to push the data to Rachio (using a $10 Pi Zero W as a Bluetooth to WiFi bridge). Previously, people talked about an API that allows moisture percent to be pushed to Rachio, but I can’t find the details of that API. (I plan on having multiple moisture sensors to measure different zones, so hopefully the API supports that.)

Or, if IFTTT is the only option for Rachio Integration, is there a way for the Rachio to accept the moisture percent in the JSON body of the Webhooks? Currently, the only relevant Rachio slots would be “Start watering time” or “Start a zone” but neither of those accept a moisture level.

(Side note: The “GreenThumb” soil sensor seems like a good alternative sensor if it was more widely available.)

Edit: Found the API here, but I don’t see anything that would allows input of soil moisture percent: https://rachio.readme.io


#2

Business idea: Perhaps Rachio could design a “smart” thermostat that never measures temperature. Simply enter the square footage of your house and the BTU output of your furnace and it can use local weather to determine when to turn on/off the furnace by guesstimating the temperature.


#3

This is interesting @dpryan- how do you see this working in comparison to Nest?

PS- I do not believe that we currently allow soil moisture to be changed through the API.


#4

@mckynzee - I was being facetious… a sprinkler controller without a moisture sensor is like a home thermostat without a temperature sensor. Open-loop control systems suffer from offset errors, sensitivity to disturbances, and a high reliance on accurate input parameters.

It seems like many other people realize the desire for this:
@johnny2678 says “Incorporating actual soil sensors would be nirvana” in this post.

@azdavidr says “Integrating robust soil moisture sensors seems like the holy grail for smart irrigation as they complete a closed-loop system” in this post.

@franz, @plainsane, @jjpaz and others talk about integrating soil sensors extensively in this thread (Feb 2015 - June 2016).

It seems crazy (to me) that Rachio doesn’t allow the input of soil moisture data.

In the meantime, it seems like the workaround is to disable all schedules on the Rachio and write your own closed-loop control system to trigger individual watering sessions using the soil moisture data.


#5

It may seem crazy at first, but have you seen how expensive controllers with 20-30 sensors cost?


#6

@plainsane That seems like a different ballpark. I’d like to have one sensor per zone. But I guess it would be expandable to as many as you want, so long as you have wireless reach. The limit with my proposed approach is the range of the Bluetooth LE in the Xiaomi sensors (since they need to reach the wall-powered Bluetooth/Wifi Bridge). The GreenThumb sensors have built-in wifi which would give a bigger range (but the battery life is only a few months).

I guess I’m not sure what your point is?

The key issue is that Rachio doesn’t have the ability to use soil moisture data, regardless of how many sensors you have.


#7

I am skeptical, but let us know if this works. Two popular soil moisture sensors installed by contractors are the Toro Precision Soil Sensor and Rain Bird SMRT-Y soil sensor. The cheapest price I could find is $109.99 plus shipping.


#8

@robertokc:
If you’re skeptical of the Xiaomi sensor, there’s a few days worth of example data here. (His tomatoes are apparently watered twice daily, so perhaps this isn’t a good example to see variation in the soil moisture.)

If you’re skeptical of the GreenThumb sensor, there’s a 2.5 weeks of example data here.


#9

Look at soil moisture sensor studies from University of Florida, University of Arkansas and New Mexico State University. They tested several moisture sensors in turf plots, in residential installations under sound experimental design that used statistical analysis. Show me university studies with the products you mention. This is why am skeptical.


#10

Here’s some interactive data that I collected from the Xiaomi/MiFlora sensors over the past two weeks:

I have two sensors, a couple feet apart. Initially, the moisture readings were close together, but then I “reseated” them to make sure I wouldn’t hit them with the lawnmower. I haven’t touched the sensors since. I was able to mow over them with my blade set to ~3.25", but I’d be nervous to cut lower than that.

Since the moisture reading is done by measuring capacitance, I suppose the exact reading depends on how tightly the soil is compacted around the sensors, so I might’ve disturbed that a bit when reseating the sensors. It still seems to give decent relative measurements, despite the offset between the sensors.

Note that there was predicted rain on Aug 20 which caused my 5am watering schedule to skip that day, even though I didn’t end up getting much precipitation.


#11

I have never heard of this company. They have never been an exhibitor at the Irrigation Association show. I think if you want soil sensors a better choice is to buy a Toro Evolution controller with smartlogic wifi and add 2 Toro Precision Soil Sensors with the controller.
Your best bet is to ditch the China or Jap sensors and use Rachio with flex daily. Trust its accuracy. We don’t need extra sensors with Rachio.


#12

Bluetooth LE range would be limited. You’d be better off with ZWave, but you’re talking a lot of battery changes… at least at the start of every season.


#13

The Toro Precision Soil Sensor uses radio frequency. It uses 3 AA batteries. Battery life is long. Toro’s soil sensor technology was originally developed for its golf irrigation division. But I don’t see the need for a soil moisture sensor with Rachio. It is better suited to use it with standard stand-alone controllers or with Toro’s SmrtLogic product. Still, Rachio is better than this.


#14

@void64, I agree the range isn’t ideal, but my solution was to find an unused outlet to hide my BLE to WiFi bridge. It’s the trade-off of having small, low-power sensors… they run for a year on a CR2032 watch battery.

With my setup, I have coverage over my (small) backyard using the Pi Zero W as a bridge. I’m considering converting the outlet to one with built-in USB ports so that I can still use the AC plugs if I wanted to.


#15

i hope you can get something working at a decent cost, i would love to drive things off of moisture sensors…


#16

It cost me $10 (+$6 S/H) for the Pi Zero W from Sparkfun, and $12/each for the moisture sensors (from “paska2016” on ebay).
So $40 for my 2 sensor setup.

The issue is that there’s no way to integrate the data into Rachio (aside from turning off all of your schedules and just commanding individual watering sessions).
That’s why I created this thread…


#17

yea, you are stuck commanding your own zones.

i personally need 14 of these sensors so that is not too bad in cost.


#18

Why don’t you use Rachio as it is intended? Why the soil sensors? Rachio has been very accurate for Oklahoma City conditions.


#19

I’d give my left arm and right big toe for an Rachio API integration to the EDYN soil sensor.


#20

I was interested in Edyn with all of the recent posts here about it, but it doesn’t seem “available” anywhere… :frowning: