Hello- on the insights graph, the date x axis marks appear to intersect at the end point for a day. For example, in my screenshot 7/10 displays as the axis mark, but detail date above is actually 7/11. Is this built as intended? It always throws me off when reviewing identification of watering/rain events. I had a rain event on 7/10, but visually in the graph it looks like it occurred on 7/9. Also, is there a lag period for when watering/rain events appear in the GeoDrops app? I always seem to see watering events display/update a day or so later than when they occurred, even though the app displays more recent last sent by device or lasted updated by app timestamps.
Hi @Ziggy ,
Apologies for the late response. I have documented this response internally. Our engineering team will have more to say soon.
Thanks for reporting this,
Stanley
Marketing PM
Team GeoDrops
Hi @Ziggy ,
The date label alignment on the x-axis is a known issue, and it can definitely be confusing. This will be addressed in a future update to make things more intuitive.
Regarding the delay in watering showing, there can occasionally be a short lag (typically less than a day) depending on when the data is received and processed. However, it shouldn’t consistently take more than a day. If you’re still noticing delays beyond that, definitely let us know.
Thanks again for the feedback!
Leon,
Engineering
Hi @Ziggy ,
Just checking in to see if the problem is resolved before I close off this thread.
Stanley
Marketing PM
Team GeoDrops
Hi @Ziggy ,
Can you post the information here that you sent via DM or are you still having problems with this?
Stanley
Marketing PM
Team GeoDrops
Can the team share what logic is utilized to determine when a watering event is displayed? Is it a minimal threshold change for increased moisture content? Attached you will see what the app shows for my device vs. the actual irrigation event data from my controller. The app doesn’t display all events (e.g 8/15).
Thanks.
Hi @Ziggy ,
Just wanted to give you a quick update. Engineering is looking into this. However, it will likely take us some time. We’ll provide another update as soon as we investigated further.
Thanks for your support and patience,
Stanley
Marketing PM
Team GeoDrops
Hi Ziggy,
The “watering detection” is based on an AI self-learning algorithm. To explain this in simple terms, AI tries to learn about your soil’s natural response to watering, and tries to detect “dominant watering events”. So it’s not a simple moisture threshold, because there are too many different soil types, and a “simple threshold” doesn’t work if we need GeoDrops to work across both sandy soil and clay soil.
For your graph, it looks like the location that the GeoDrops device is installed in is actually NOT getting enough water! And so the large watering event (could it be a rain event)? Brought the soil moisture high up, and the AI has decided that “this large watering event on Aug 15~16” is the new norm for the baseline of a “watering event”.
If AI detects doesn’t detect such large watering event for a a while, it’ll slowly revert back to allow a smaller watering increase to be counted as “watering events”.
But I think a bigger take away is that you may need to water more in the area where this GeoDrops device is installed, since it’s clearly not retaining enough water based on your daily watering. 
Please let me know if this helps?
Thanks,
Lawrence
Hi @homedigy-lawrence, thanks for the additional insight on measurement. Our area experienced a significant drought with not much rainfall. Despite my best effort to keep the lawn watered, I was not able to keep up with the hot temperatures.
The event you see on Aug 15-16 is a reset of max moisture, the device prompted me to do so and I saturated the soil near sensor. The last measurable rainfall we encountered was on Aug 21. Thanks for helping me learn how the device is working. It helps me adapt the information and determine action I can take.
Hi @ziggy,
I just want to check if your problem is resolved, or if you need more help with this.
Stanley
Marketing PM
Team GeoDrops
We can close. I’ll send a separate note if I continue to have odd results.
1 Like