I know that this project has been dormant for a while, but 2020 and all. Anyway, I am back at it and finished the research and design portion for the soil sensors.
That lead to a simple design that can use an ESP8266 (or Arduino, etc.). The difficulty of this portion was that a simple resistive probe, like those commonly on Amazon, Ebay, Banggood, etc., was not accurate enough to use for agricultural quality soil measurements. This led to the packed gypsum Watermark probes. These are reasonably priced and are widely used by university research programs throughout the US. For that reason, proper calibration information is easily obtained.
The more difficult part for me has been deciphering the technical math and determining how that relates to real world watering requirements. This obviously varies significantly depending on the type of plants and soil composition. However, the relevant research papers will get our sensor system close enough to create a usable starting point. Then if the user decides to take field calibration measurements for their specific situation, this sensor system should be able to be adjusted to that actual measurement.
Obviously it will not replace good agri-management, but it should be a useful tool for that job.
Once the prototype is tested and operational, I will post the relevant soil and Soil Water Potential (SWP) and other data to my github (txNgineer). All of my software will be up there as well.