Blog
Two plots grown identically, and one yields less every year
Precision agriculture does not start with drones or satellites. It starts with knowing where the water actually reaches.
Published
A durian grower in the east has two adjoining plots. Same age of tree, same variety, same fertiliser, same irrigation system. Every year one yields visibly less than the other.
He knows they differ. He does not know why, and the explanation he has always been given is that the soil is better over there — which is not wrong and does not tell him what to do.
Put soil moisture sensors at several points in both plots for a season and the picture sharpens. The weaker plot has spots where water sits longer after rain, and points at the end of a row that dry faster than anywhere else because pressure does not reach them. Neither is visible to someone walking the rows.
Where precision agriculture actually begins
The phrase usually arrives with pictures of drones and satellite imagery, which genuinely help at large scale. For a mid-sized commercial farm in Thailand, the fastest return is almost always water.
Water is the most controllable input, the most directly measurable, and the clearest cost — in pump electricity and in yield foregone.
Soil moisture sensors are inexpensive and not hard to install. What is hard is placing them at the right points and depths, and turning readings into decisions rather than a graph on a screen.
What the data answers
When to irrigate, rather than every how many days. A fixed schedule is a guess repeated. Soil after twenty millimetres of rain and soil after five are days apart. Irrigating on a schedule means over-watering half the time and under-watering the other half.
Where water is not arriving. A drip system that was well designed on installation day degrades with clogging and falling pressure. A point that is anomalously dry in the data is a point where somebody should check the line.
How moisture relates to yield. After two or three seasons it becomes visible which growth stages are most sensitive on that particular plot — knowledge specific to that land, not general advice from a manual.
Constraints to accept
A farm is brutal on equipment. Humidity, sun, dust, animals and machinery destroy hardware not built for it within months.
Signal is another matter. Plots far from a tower need a different transport, and sensors have to buffer readings locally when they cannot send, then backfill when the signal returns.
And batteries. A sensor needing a change every three months across a hundred points is a system that will be abandoned inside a year.
Start small and measure honestly
One plot, ten measurement points, one season, compared against a neighbouring plot managed the usual way.
That comparison needs no advanced statistics to tell you whether it was worth it, and it produces exactly the information needed to decide about expanding next season.
What GIPSIC does here
We design IoT hardware, write the firmware, and build the cloud platform behind it, with our own installation teams across several provinces. Agricultural work has to be designed around durability and power draw from the beginning, not once devices start failing.
What we usually propose is measuring first and automating later. A system opening and closing valves on data nobody has yet verified can do damage faster than it does good.
If you would like to talk about your land, get in touch.
About the author
Film — Wisit. A businessman who still does his own BA work more often than he probably should, and writes a fair bit of code, front and back. Runs two or three small businesses. Follows technology and business obsessively, in Thailand and everywhere else. Off the clock: physics, astronomy, DIY, and anything to do with networks. Music always on, though he cannot sing. Plays instruments anyway, badly. Plays a lot of sport, racket sports above all. Not much of a traveller by himself, but happy to take Mint anywhere in the world. A man who fears — sorry, loves — his wife. One flaw: he barely touches video games.
Written with Claude Opus 5