Agriculture Irrigation Management Software: How to Structure a Smart Farm

Complete smart agriculture irrigation system from pump station to field valves

Two questions come up again and again from growers moving to irrigation automation, and they’re really the same question asked from two ends.

The first is what does a complete smart irrigation system actually consist of? — because catalogues sell valves, sensors, and gateways as separate items, and it isn’t obvious how they add up to something that runs a farm.

The second arrives later, usually from someone already running equipment: how do I keep this organised once there are hundreds of devices? A recent question from a vegetable grower in Austria put it precisely — could the software sort by crop, or by region?

This guide answers both, in that order. First the physical system, from the water source outward. Then the structure your irrigation management software needs to impose on top of it — the part almost nobody explains, and the part that decides whether a large farm stays manageable.

Part One: The Physical System, Source to Field

Start at the water and work outward. Every automated agricultural irrigation system, whatever its size, is the same chain of components — and understanding that chain is the first step toward precision agriculture irrigation, because you can’t apply water precisely until you know what each part of the system controls.

Working principle diagram of a LoRaWAN smart irrigation system showing pump, filter, master valve, field valves, gateway and cloud
The full chain — from water source and pump station, through the main line, to individually controlled valves in the field.

1. Water source and pump — the head of the system

Everything downstream depends on what the pump can deliver. In an automated system the pump isn’t operated by hand at the pump house: a pump controller is added to the existing starter arrangement, so the pump can be started and stopped remotely and, more importantly, sequenced together with the valves. The controller signals the starter you already have — it doesn’t replace your electrical installation.

2. Filtration — before anything else in the line

Filters sit immediately after the pump, protecting everything downstream. This is standard practice in any irrigation system and it matters more once emitters are involved: drippers are far less tolerant of particles than an open pipe.

3. The master valve — one switch for the whole main line

A single valve on the main line, downstream of filtration, controlling whether the system is live at all. It’s what lets you isolate the entire network with one command rather than closing valves one by one.

4. Distribution valves — where control actually happens

The main line branches, and each branch gets its own smart agriculture valve. These are the working units of the system: each one opens, closes, and holds a set opening percentage on command, serving its own section of drip line or sprinklers. Because each is solar-powered and wireless, adding one doesn’t mean trenching a cable back to a control cabinet.

5. Measurement — attached to valves, not standing alone

This is the layer that turns remote control into genuine irrigation monitoring and control. A flow meter records how much water passes its valve. A pressure sensor reports pressure at its valve and raises an alarm when it moves outside the expected range — the signature of a burst or a blockage. A soil sensor reports moisture in the root zone, which is what lets irrigation respond to conditions rather than to a clock. All three are optional, and this next point matters for the software section below: each one belongs to a specific valve. There’s no such thing as a free-floating sensor reading.

6. Connectivity — the layer that makes it a system

Field devices need a path to the cloud. On most large sites that’s a LoRaWAN gateway serving the whole farm, with the gateway itself reaching the internet through a SIM card or an ethernet connection. On smaller installations each valve carries its own 4G SIM instead. (For which suits your site, see 4G vs LoRaWAN.)

7. Cloud and app — where you actually work

Commands, schedules, measurements, and alarms all pass through the platform to your phone or computer. This is where remote irrigation monitoring stops being a feature list and becomes the way you run the farm — and everything in Part Two happens here.

THE PATTERN — Source, pump plus controller, filtration, master valve, distribution valves, optional measurement, connectivity, platform. A 5-hectare block and a 1,000-hectare operation have the same architecture. What changes is how many valves hang off it — and that’s exactly why the management structure matters.

Part Two: The Management Hierarchy

Hardware is the easy half. Ten valves you can keep in your head; three hundred you cannot. What keeps a large installation workable is a structure that never changes, no matter how much you add to it — and this is the real test of any irrigation management software: not how many features it lists, but whether its underlying model still makes sense when your farm has grown.

It has three levels.

Farm management hierarchy diagram showing farm, plots with crops, and valves with bound flow meters and pressure sensors
Three levels, and two things that are deliberately not levels: crop, and sensors.

Level 1 — Farm: the management unit

The farm is the top of the structure and the unit you operate in. One site, one farm. Each farm has its own device list and its own overview.

If you run several separate sites, you create several farms — but with an important qualification: day-to-day management always returns to a single farm. You don’t operate three sites simultaneously from one screen, because they don’t share a pump, a main line, or a schedule. The multi-farm capability exists so that separate operations stay separate, not so they can be merged into one blur.

Level 2 — Plot: the smallest thing you irrigate as a unit

A farm is not one piece of land. It divides into plots — and the working definition is precise: a plot is the smallest area you want to irrigate as a single unit. Not the smallest area you own, and not necessarily one field. If two halves of a field are always watered together, they’re one plot. If they’re watered separately, they’re two.

This definition is doing real work, because it’s what makes the whole structure line up with how you actually farm rather than with how the land is drawn on a map.

Level 3 — Valve: exactly one plot each

Under each plot sit its valves. A small plot may need one; a larger plot will have several sharing the irrigation work. The rule that keeps this unambiguous: every valve belongs to exactly one plot.

So there is never a question about which area a valve is watering — and, read in the other direction, never a question about which valves to command when you want to irrigate a given plot.

What is deliberately not a level: flow meters and pressure sensors

This is where a lot of software gets it wrong, listing sensors as separate devices alongside valves. In this structure they are bound inside the valve they measure.

The reason is logical, not cosmetic. A flow meter measures what passes through its valve. A pressure sensor reports pressure at its valve. Neither reading means anything detached from the valve it came from. Keeping them inside the valve means every measurement has exactly one valve, one plot, and one crop behind it — there is no such thing as an orphaned reading you have to trace back to its source.

Why Crop Is a Property, Not a Level

Here’s the design decision that matters most over time, and the one that answered our Austrian customer’s question.

Crop is stored as an attribute of the plot, alongside things like its section of the farm and its area. It is not a structural level. The reason is a single rule:

THE PLOT IS FIXED. THE CROP IS A VARIABLE. Boundaries, pipework, and valves stay where they are for years. What grows on that ground changes every season with your rotation. Structure should follow the thing that doesn’t move.

Consider the alternative. If crop were a level in the hierarchy — farm, then crop, then plot, then valve — then every rotation would tear the structure apart. Move carrots from Plot 1 to Plot 4 and you’d be rebuilding the relationships between crops, plots and devices, season after season, purely because the software insisted on a level that doesn’t correspond to anything physical.

As a property, a rotation is one field edit. Plot 1 grew carrots; now it grows potatoes. The pipework, the valves, the sensors and every historical measurement stay exactly as they are.

Which is the honest answer to “can it sort by crop or by region?”: yes, and precisely because neither one is a level. Both are properties, and properties are what you filter on.

Setting Your Farm Up in the App

The structure above translates directly into setup. Five steps, done once.

Step 1 — Create the farm

Create a farm and name it. If you operate several separate sites, create one farm per site — but set them up as genuinely separate entities, because that’s how you’ll manage them.

IOTMGR app screen showing a new farm being created and named, with the responsible staff assigned
Creating the farm — name it for the site, and assign the people responsible for it.

Step 2 — Assign who is responsible

Add the people responsible for that farm. On an operation with several sites or several staff, this is what stops alarms going to someone with no ability to act on them.

Step 3 — Create the plots

Within the farm, create each plot, using the definition above — the smallest area you irrigate as a unit. Record the crop for each plot, along with its area and its section of the farm. These are the properties you’ll filter on later, so it’s worth being consistent about them now.

IOTMGR app screen showing a plot being created inside a farm, with its crop and section recorded
Creating a plot — name it by section, then record the crop and area as properties of that plot.

Step 4 — Assign valves to their plots

Add the valves that irrigate each plot. Where a valve carries a flow meter or a pressure sensor, those are bound to the valve itself, so they arrive in the structure automatically rather than needing to be placed separately.

Step 5 — Name every valve properly

The step that looks trivial and isn’t. See below.

Why Valve Names Decide Your Response Time

Naming valves feels like housekeeping. On a large farm it’s operational infrastructure, and here’s why.

At two in the afternoon a pressure alarm arrives on your phone. What happens next depends entirely on what that alarm says.

If the valve is called “Valve 47”, someone now has to work out where Valve 47 is. Maybe there’s a list somewhere. Maybe the person on site today wasn’t the person who installed it. On 1,000 hectares, that’s an afternoon.

If the valve is called “North Section – Plot 3 – Carrots – Valve 2”, someone is already driving there.

WHAT THE STRUCTURE BUYS YOU — Because every sensor sits inside exactly one valve, and every valve sits inside exactly one plot, an alarm doesn’t just tell you something is wrong. It resolves automatically to a single device, in a single plot, growing a known crop, in a known section of the farm. The hierarchy turns an alert into an address.

Two practical habits make this work:

  • Name by position, not by serial number. The device identifier already exists and is engraved on the unit; the name field is for humans in a hurry.
  • Use one convention across the whole farm, decided before installation rather than invented valve by valve. Consistency is what makes names sortable and searchable later.

Worked Example: Structuring a 1,000-Hectare Farm

Abstract structure is easy to agree with and hard to apply. So here is the whole thing done concretely, on an operation of 1,000 hectares — and the first question decides everything that follows.

ASK THIS FIRST — Is your 1,000 hectares one continuous block, or is it several separate pieces spread across a wider area? Continuous means one farm. Scattered means several farms. Everything below branches from that answer.

Case 1: 1,000 continuous hectares → one farm

If the land is contiguous — one site, one operation — then it is one farm, no matter how large. Size is not what splits a farm; geography is.

Step 1. Create the farm. One farm, named for the site, with the people responsible for it assigned to it.

Step 2. Create the plots, and name them by where they are. Divide the farm according to how you actually irrigate, and build the location into the name from the start: North Section – Plot 1, North Section – Plot 2, South Section – Plot 1, and so on. A name that carries a location is worth far more later than a name that carries only a number.

Step 3. Record the crop on each plot. Potatoes on one, carrots on the next, onions, cabbage, peas, spinach. This is the property you’ll filter and group by — and the one you’ll edit next season without touching anything else.

Step 4. Add the valves, and name them after their plot. This is the step that pays for itself repeatedly. The valves on North Section – Plot 1 become NS1-01, NS1-02, NS1-03, NS1-04. The valves on South Section – Plot 2 become SS2-01 onward. The name of the valve tells you the plot it belongs to before you’ve looked anything up.

Structure for a continuous 1,000 hectare farm: one farm containing six plots, each with its crop and a block of valves named after the plot
A continuous 1,000 ha operation as one farm — plots named by section, valves named after their plot.

Case 2: 1,000 scattered hectares → several farms

Now the other situation, which is at least as common: the same 1,000 hectares, but split across five separate pieces of land, perhaps 30 km apart.

Forcing that into one farm produces a device list that mixes five unrelated sites together. Instead, create five farms — one per site: Farm 1, Farm 2, Farm 3, Farm 4, Farm 5, each named for its location so the name means something on the phone.

And then — this is the important part — you go back to Case 1. Open one farm, and build its plots exactly as above: named by section, crop recorded, valves named after their plots. Repeat for each farm.

Structure for a scattered 1,000 hectare operation: five separate farms, each then divided into its own named plots
Scattered land becomes five farms — and inside each one, the same plot-and-valve structure repeats.

The reason this works is that the two levels answer different questions. The farm answers “which site am I working on?” The plot answers “which piece of that site am I watering?” Trying to make one level do both jobs is what produces the unmanageable device list.

Frequently Asked Questions

What are the components of a smart agriculture irrigation system?

Working from the water source outward: a pump with a remote controller, filtration, a master valve controlling the whole main line, and distribution valves serving each part of the field. Flow meters and pressure sensors can be bound to individual valves for measurement and alarms. Connecting it all is either a LoRaWAN gateway serving the whole site or 4G in each valve, which links the field equipment to the cloud platform and the app you operate it from.

How should a farm be organised in irrigation management software?

In three levels. The farm is the top management unit — one site, one farm. Each farm divides into plots, where a plot is the smallest area you want to irrigate as a unit. Each plot contains its valves, and every valve belongs to exactly one plot. Crop is stored as a property of the plot rather than as a level of its own, and flow meters and pressure sensors are bound inside the valve they measure rather than listed separately.

Why is crop a property of the plot instead of a separate level?

Because the plot is fixed and the crop is a variable. Boundaries, pipework and valves stay in place for years, while what grows there changes with rotation each season. If crop were a structural level, every rotation would mean rebuilding the relationships between plots and devices. As a property, a rotation is a single field edit and the hardware structure is untouched — and you can still filter and group by it, for example selecting every plot growing carrots.

When should I create a second farm instead of another plot?

If two areas are geographically separate and you think of them as separate operations, set them up as two farms, each with its own device list and overview. If they belong to the same operation and you simply want to irrigate them independently, make them two plots within one farm. Day-to-day management always returns to a single farm, because the farm is the unit you operate.

How do you keep track of hundreds of valves across a large farm?

Through the hierarchy and disciplined naming. Because every valve sits inside exactly one plot, and every sensor sits inside exactly one valve, any alarm resolves to a single device in a single plot growing a known crop. Naming valves by their position — rather than by serial number — is what turns an alert into a location someone can drive to, which is the difference between a five-minute fix and an afternoon of searching on a large site.

Mapping your own farm?

The hardware in this guide:
>>> Ball Valves · Butterfly Valves · Flow Meters · Pressure Sensors · LoRaWAN Gateway · Pump Control

Related reading:
>>> 4G vs LoRaWAN · App control functions · Watch one full irrigation cycle end to end · The Complete Guide to Smart Irrigation Valves

We’ll set the structure up for you.
>>> Send us your plot list — plot names, crop and area for each, and how many valves per plot — and we’ll prepare the structure for your account so you can see it working with your own data. Or tap the WhatsApp button in the corner of any page.


AriValve builds solar-powered, motorized wireless smart irrigation valves and the platform that manages them, deployed across 30+ countries. This guide came out of a real customer question about organising a multi-plot vegetable operation.

Leave a Comment

Your email address will not be published. Required fields are marked *