The challenge
Automated irrigation is reshaping agriculture, but the starting point on most farms is surprisingly blunt: a fixed schedule. Blocks get watered on a routine that doesn't know it rained overnight, that tomorrow will be 38°C, or that one block sits on sandier soil and drains twice as fast as its neighbour. The result is waste on three fronts at once — water applied to ground that's already wet, electricity burned pumping when the crop doesn't need it, and labour spent driving the farm to manage valves by hand. And underneath it all sits the real risk: crops stressed by getting too much or too little water, which quietly costs yield and quality.
The core problem: the irrigation decision is being made by a calendar, not by the crop. A schedule can't see the weather, the soil or the moisture already in the ground — so it systematically over- and under-waters, wasting inputs and stressing the crop at the same time.
Our approach
Sensor-driven irrigation moves the decision from the calendar to live data:
- Soil-moisture sensing per block — watering follows the actual moisture in the ground, at representative points and depths, so each block gets what it needs and no more.
- Weather-aware scheduling — rainfall and evapotranspiration feed the schedule, so the system holds back after rain and anticipates a hot, thirsty day.
- Automated pump and valve control — pumps run only when the crop needs water, and scheduling can shift pumping into cheaper tariff windows and around load shedding rather than being defeated by it.
- Off-grid sensing on remote blocks — solar-powered nodes report over long-range, low-power networks (LoRaWAN / NB-IoT), so distance from power and signal is no barrier, and data buffers through outages.
- One farm dashboard — moisture, pump runs and faults visible at a glance from anywhere, replacing the drive-around.