Irrigation automation is the management of a hydraulic process — the soil water balance — using control engineering methods. Standards such as ASABE EP596 (irrigation controller terminology, 2014) and ISO 15886 split controller behaviour into two main approaches: scheduling-based and sensor-driven. In practice, modern systems use a hybrid combination of the two.

Control architectures

The principal closed- and open-loop control strategies applied in irrigation:

  • Open loop (scheduling): based on time and duration. There is no feedback; this is what most garden timers use.
  • On/off (bang-bang) control: irrigation is started when soil moisture falls below a lower threshold and stopped when it reaches the upper threshold. Hysteresis prevents oscillation.
  • Set-point control: a target moisture or water potential (typically in the −20 to −60 kPa range) is defined; PI/PID-style continuous controllers adjust flow or duration.
  • Model-predictive control (MPC): the soil water-balance model and short-term weather forecast are used to optimise water demand over the next 24–72 hours. The FAO-56 Penman-Monteith (1998) ETo estimate is a frequent input.
  • Fuzzy logic: uncertain inputs (very dry, mild, windy) are evaluated using linguistic rules; robust against sensor noise.

Rule-based approach

Rule-based control combines time, sensor thresholds and weather variables using "if–then–else" logic. Its advantages are transparency and verifiability by the operator; the drawback is the risk of combinatorial explosion as the number of rules grows. Typical inputs: absolute time (hour, day, season), soil moisture (% volumetric water content or kPa matric potential), reference evapotranspiration (ETo, mm/day), rainfall forecast, wind speed (winds above 4 m/s degrade droplet distribution) and frost alarm.

Hierarchical zone management

In multi-zone systems, irrigation is executed sequentially rather than in parallel because resources (pump, line pressure) are limited. The hierarchy is typically three levels deep:

  • Site: hydraulic capacity, master valve, water source.
  • Program / schedule: a scheduled task covering one or more zones.
  • Zone / station: a single solenoid valve and the irrigators connected to it.

Overlapping programs are placed in a queue and resolved through priority, delay and limits on simultaneous operation (max simultaneous stations). This logic is outlined in ASABE EP405.1 and ICC/ASABE 802.

Practical threshold: a rain sensor triggers on rainfall above 4 mm; this value partially offsets daily ETo for most turf cover (5–7 mm/day in a Mediterranean summer). Rainfall below 4 mm is subtracted from the daily water budget but does not cancel the program entirely.

Adaptive adjustment

The seasonal adjust method scales the base program monthly or weekly by the ETo ratio. More advanced systems apply ET-based irrigation: net water requirement in litres is calculated from daily ETo × crop coefficient (Kc) × area (m²) − effective rainfall. Crop coefficients are taken from the FAO-56 tables (e.g. turf Kc ≈ 0.85; mature deciduous tree ≈ 0.70).

Manual override

Automatic control architectures always provide a manual override layer. Operator commands run with the highest temporary priority; once their duration expires, control returns to automatic. In the industrial-control literature this pattern is referred to as "operator-in-the-loop" and is recommended in the ISA-101 HMI guideline.