I’ve been searching quite a while to see if there is a blew print around but in vain.
I have a Gardena Sileno Smart which can by default only run on a time schedule, but being a HA user, this is not enough for me.
What I am trying to achieve is to make it mow only when the grass is dry. So it has not rained and there is now dew (best would be if the evaporation is being calculated so even when there was dew/rain, the grass is being marked as dry again when all water evaporated). I have tried to calculate dew amount and evaporation with some simplified formzlars, but failed.
Is there any hardware, template or blueprint which is taking care of this?
Quite a few smart mowers have a custom component written by someone, there’s no reason you can’t smarten your mower up quite a bit. It all starts with what your mower can do in HA manually (can you start it, dock it, etc?), if you have those controls now then you don’t need a blueprint, you simply need some simple automations combined with perhaps some sensor templates to do the job.
For instance you can create a sensor that keeps track of rain but also of sprinklers, to know how wet your grass basically is, that sensor could just return a true/false for if mowing is a go or no go at any given time. Make it smarter by adding attributes to your sensor to calculate how many hours it’s been and then make an automation to trigger on “when sensor reports dry for 10 hours and it’s still daylight and it’s Mon-Wed-Fri then go mow”. That’s super easy to do.
I have a Gardena Sileno which is integrated via the HACS integration.
I currently track the rain via the openweather integration, as well as Temperatur and UV Index. I also have sensor that calculates whether the conditions for dew building are met.
The struggle I have is to create a sensor that tells me if any possible moisture (being rain or dew) has been evaporated taking the temperature and relative humidity into consideration.
First I thought I can do this via a “time since last rain” or “time after sunrise” sensor but there would be too much variation due to temperature and humidity.
Basically I want the mover to have time to mow between sunrise and sunset for a given amount of time (easy to do) but place blockers for wet grass, too high temperature and too high UV-Index which are checked every thirty minutes.
So what you are looking for is evapotranspiration (ET) calculations, which are pretty tricky to write. I’ve written them, there are also a couple plugins here that can do it, but there are so many factors that rely on more than OpenWeatherMap can provide, so you end up bending the math to fix the holes in the data. For me I wanted this to know when to water outdoor plants and grasses based on the evapotranspiration algorithms.
Fortunately my issue ended being resolved to “good enough” with using the Rachio sprinkler since it actually uses evapotranspiration in knowing when to water and for how long. Applying this to a lawn mower, while really scientific and nerdy, is probably overkill. You’ll end up pulling your hair out dialing it in when you can do very anecdotal calculations based on how much time has passed since there was rain or even putting a sensor outside to measure the water saturation of the soil - none of which really tells you whether the blades of grass are still too moist to mow.
But, since mowing generally only happens during growing season, which tends to be warm, you could do basic calculations of approximately how much time is needed in direct sun at, say, a 70 degree F day, before mowing is OK for you.
To give you an idea of how tricky evapotranspiration can be, it needs to know the temp, humidity, dew point, heat index, UV score and slope of every zone you mow in, the soil composition of every zone you mow in, the precise water on your property (this is one of several ways that OpenWeatherMap will fail you) and so forth. It’s tricky stuff! A blueprint could take you only so far, much of this data is hard to come by unless you have a local weather station (I use a Tempest myself) and do soil composition tests (which I’ve done) and still have sensors in the ground.
I already thought that this might be a bit of an overkill but it somehow interested me if there a ways to fix it. I thought I could scale it down to one square meter since my garden is quite small. I have a sensor thats in the ground to measure the moisture.
So you would suggest to just add a delay to the last rain and the sensor that tells me if conditions for dew have been met rather than trying to overcomplicate things?
I wonder how many times paying the local teenager or even a professional mowing company it takes to pay for a robot mower. I’ve been wanting one for years but the cost of them, in addition to the overwhelming negative experiences I’ve read about have convinced me to just build my own out of someone’s Craigslist used electric mower (one day).
The Husqvarna 546 manages up to 48.000 square meters, but at a price at around 35k USD you can get you neighbor to cut your lawn 700 times before it’s worth it. Or every week year round for 13,5 years.
I’ve really thought this out, an old used electric mower, put on some turning wheels, connect to a waterproof ESP that can basically do the same thing. Probably not as graceful, unless you want to bury cables like you do with commercial mowers, you need ultra accurate GPS and wifi but it is doable, a fun project when I have a couple hundred hours to spare. Either that or build the open source version, but even that one sets you back quite a bit.
3500USD for a basic Husqvarna sounds like a very high price - Here even with a horrible exchange rate to USD we are around 2.2k USD for a Husqvarna 405X (X is the top range on the private market).
Still quite expensive, but running on year 2 and 3 on our mowers it starts making sense For my dad which cannot mow himself anymore it’s really a no-brainer.
I had a colleague with plans of making a VR mower completely DIY, his plan was to use a VR headset and mow the lawn from the couch, building the mower with RC parts. Fun project but I don’t think it came to reality.
