I read that post for the first time today, thanks for sharing. But to me this is an example of where rachio could with a minimum of effort do this a bit better. From what i gather smart zone (which by the way does not appear anywhere in the app -what is the point of branding something if it is not visible in the app?) is putting in what rachio considers to be defaults based on commmunity input rather than theoretical models.
I think that is a good thing -but how do i know which one it is using? And if its not a new capability, just another method for filling in the default settings, how does the install base know it is there? I suspect it provided the defaults for me, but nowhere do i know if these have been adjusted (or not) for my street/city/region.
I applaud getting more correct defaults, but it by itself doesn’t reduce complexity nor did rachio do much to advertise to inform me that they have a way to get the app to setup the zone correctly for my yard faster and with fewer errors. I think a popup saying "90% of users who explicitly programmed in settings for their lawn in your neighborhood used these settings would be so useful.
Someone asked me to choose flex monthly instead of daily. I think this is a false narrative. The choice should not be between efficiency and ease of use. I want to have an efficient watering system. I do not wish to reduce efficiency to gain simplicity -or if i do that then i think that efficiency loss should be quantified before i make it.
But in general I think it would be wise to implement a system that lets me gain that efficiency without introducing complexity. Or if it requires me to accept more complexity at least articulate how much more efficiency gains i get from that complexity.
But alas, monthly or daily schedules seems to have little impact on how to best select settings like crop coefficency, root depths, depleted water, nozzle types, efficiencies. Those are all transposed across all schedules. Imho this quickly becomes a false choice - it doesn’t particularly reduce complexity and its not clear to me that any efficiency was lost or gained either way.
And just so i am clear. I am writing this feedback to show that i care, not to be overly harsh or to be a whiner. i hope that my feedback and criticism helps build a better platform for all.
Nest uses leaves for community feedback based on individual history and group history. I find nest remarkably easy to use, i used the ecobee for the remote sensors but it had sufficient more complexity vs. nest that i never could get my family to love the thermostat. Once nest had remote sensors out my ecobee went to donation. To me Rachio faces an variation of that, they are at the brink of greatness but its easy to loose that vision when faced with a ton of capabilities and opportunities.
To me the brand promise of a rachio is less effort to maintain my garden while gaining water efficiencies.
Lets make sure that this promise is delivered!
Final point, for ML, if this approach is going to adjust watering schedule make sure that you have a really solid regression testing for your data sets so that the predictor is correct for a number of well known data sets. If my garden gets destroyed because of algorithmic regression then i’ll get seriously upset. I’ve seen more than a fair share of ML approaches going sideways as data changed and inputs created unforeseen predictors and models. Be super careful…