Farm monitoring system projects rarely need to start with a new platform. Most farms already own more sensing kit than they realise: a weather station on the workshop roof, temperature cables in the grain store, a bulk tank controller with an alarm relay, an EID stick reader, telematics in the tractors and a phone in every pocket. The problem is that each device reports to its own app, its own panel or nobody at all. The information exists, but nobody sees it at 3am when a store starts heating or a trough runs dry.
This guide shows how to turn that scattered kit into one working farm monitoring system. It covers how to audit what you own, get data out of each device, connect it over patchy rural signal, choose software that fits the farm, set alerts people act on, and budget for the few pieces you genuinely need to buy.
It is written for farmers, farm managers and the advisers who support them. If you are building a farm software product rather than a farm setup, our agritech software development guide covers that job, and our earlier look at IoT for smart agriculture sets out the wider technology.
Every threshold and price reference below comes from a named public source, such as AHDB grain storage guidance or the Farming Equipment and Technology Fund 2026 item lists, or is arithmetic you can check. Where a figure is an assumption, it says so.
Table of contents
- What a Farm Monitoring System Actually Needs to Do
- Start Your Farm Monitoring System With an Audit of What You Own
- Four Ways to Get Data Into a Farm Monitoring System
- A Farm Monitoring System Architecture That Suits Most Farms
- Connectivity: Getting Farm Monitoring System Readings Back From the Field
- Choosing Software for a Farm Monitoring System
- Farm Monitoring System Alerts People Will Not Ignore
- How to Build a Farm Monitoring System in Five Stages
- Worked Example: A Farm Monitoring System for a Mixed Farm
- What a Farm Monitoring System Costs, and What FETF 2026 Paid For
- Security and Data Ownership in a Farm Monitoring System
- Common Farm Monitoring System Mistakes
- When to Bring In Help
- Farm Monitoring System FAQs
- References and Further Reading
What a Farm Monitoring System Actually Needs to Do
A farm monitoring system has three jobs: watch the conditions that matter, keep a record of them, and tell the right person when something drifts out of range. Everything else, from colourful dashboards to yield forecasts, is optional until those three work reliably.
Watch the few things that cost money when they go wrong
Start from losses, not from sensors. On most farms the expensive surprises are familiar: grain heating in store, milk not cooling in the bulk tank, water running out in a shed or a field trough, frost on a high-value crop, a power cut that stops fans or pumps, and a gate left open. A farm monitoring system earns its keep by catching those early. A sensor that measures something nobody would act on is just another battery to change.
Keep a record you can use later
Records matter for more than curiosity. Grain buyers, assurance schemes, vets and insurers all ask what happened and when. A time-stamped log of store temperatures or milk temperatures answers that question without anyone relying on memory. It also lets you judge whether a change, such as a new fan routine, actually worked.
Alert a person, not a screen
Nobody watches a dashboard in the middle of lambing or harvest. The part of a farm monitoring system that prevents losses is the alert that reaches a phone, reaches the right phone, and says clearly what is wrong. Dashboards help you review a week; alerts save the night.
What it should not try to replace
A farm monitoring system sits alongside the farm management software you already use for records, compliance and accounts. It should not become a second place to enter spray records or animal treatments. Keep those in the system of record you trust, and let the farm monitoring system handle live readings and alarms. Trying to merge the two is the fastest way to turn a weekend project into a stalled software build.
Start Your Farm Monitoring System With an Audit of What You Own
The cheapest farm monitoring system is the one built mostly from kit you have already paid for. Walk the farm with a notebook, ideally with the person who fixes things, and list every device that measures anything, where its readings go today, and how you could get them out.
Weather station and rain gauge
Many farms bought a weather station for spraying decisions or disease models. Many modern stations upload to a vendor cloud. Davis stations, for example, report to WeatherLink, and the WeatherLink v2 API gives access to station metadata and observation data for stations you own or that have been shared with you. That makes the weather station one of the easiest devices to bring into a wider farm monitoring system.
Grain store probes and controllers
Temperature cables and store controllers are often the most valuable sensors on an arable farm, and the most isolated. Many controllers show readings only on a panel inside the store or trigger a basic alarm. Check yours for a network port, a serial port, a data-logger export or a relay output. AHDB advises taking store temperatures where cooling takes longest: furthest from the fan in blowing systems, usually 0.5 m beneath the surface and centrally between ducts. Note where your existing probes actually sit, because a probe in the wrong place gives a comforting but wrong answer.
Livestock kit: EID readers, weigh crates, parlour and robot software
Livestock farms usually have data locked inside herd software: EID readers, automatic weighing, parlour or robot systems, and activity collars. These rarely need new sensors. What they need is a route out, usually an export, a report or an API that the vendor may charge for. List what each system records and whether it can already raise alerts by itself.
Tanks, pumps, meters and power supply
Bulk milk tank controllers, water header tanks, borehole pumps and the mains supply are classic blind spots. A tank controller may already have an alarm relay that switches when cooling fails, and a borehole pump may have a pressure switch. These simple on/off signals are some of the most useful inputs a farm monitoring system can have, because they tell you that something has stopped. Our guide to IoT for smart water management goes deeper on water.
Machinery telematics
Tractors, combines and telehandlers from the main manufacturers report location, hours and fault codes to the maker’s portal. That data is useful for service planning, but most of it is not urgent. Treat telematics as a source for weekly reports rather than 3am alerts, and leave it in the manufacturer’s portal unless you have a clear reason to copy it.
Cameras and the phone in your pocket
CCTV cameras on yards and calving pens already provide live images, and newer ones raise motion or line-crossing events. Camera analytics built on computer vision can flag a calving or a vehicle at night, but start with the simple events your recorder already produces. Phones matter too: they are the alert screen, the dashboard and, with the right app, a way to log readings that no sensor captures.
| Device you may already own | Data it holds | Usual way out | Question to ask |
|---|---|---|---|
| Weather station | Temperature, rainfall, wind, humidity | Vendor cloud API or local logger | Does my data plan include API access? |
| Grain store controller and cables | Store temperatures, fan status | Serial or Modbus port, logger export, alarm relay | Is there a network or serial port? |
| Bulk milk tank controller | Milk temperature, wash cycles, alarms | Alarm relay, service port, vendor app | Can the alarm relay be wired to a gateway? |
| Parlour, robot or herd software | Yields, activity, health flags | Export, scheduled report, paid API | Which alerts does it already send? |
| EID reader and weigh crate | Animal IDs, weights | Bluetooth sync to an app, CSV export | Can weights be exported on a schedule? |
| Machinery telematics | Location, hours, fault codes | Manufacturer portal and API | Who holds the account and the data rights? |
| Pumps and header tanks | Often nothing yet | Pressure switch, float switch, dry contact | Where would a level sensor fit? |
| CCTV cameras | Images, motion events | Recorder events, email or push | Can the recorder send events elsewhere? |
Most farms finish a farm monitoring system audit with a surprise in both directions: more sensors than expected, and fewer of them reporting anywhere useful. That list is the plan for your farm monitoring system.
Four Ways to Get Data Into a Farm Monitoring System
Once the audit is done, each device falls into one of four integration routes. The route decides how much work a device takes, how quickly its readings arrive, and how likely the link is to break when a vendor changes something.
Route 1: vendor cloud APIs
Where a device already uploads to the maker’s cloud, an application programming interface (API) is the tidiest route. Your farm monitoring system asks the vendor’s cloud for new readings on a schedule, usually every few minutes. The trade-off is dependence. If the vendor changes its terms, adds a fee or retires the API, the link breaks. Read the terms for data ownership and API access before building on them.
Route 2: local protocols such as Modbus and MQTT
Industrial controllers, including many store, ventilation and pump controllers, speak Modbus over a serial or network connection. A small gateway in the yard can poll those registers and publish the values. MQTT is the common language on the other side: a lightweight publish-and-subscribe protocol designed for small devices and unreliable networks. It defines three quality-of-service levels (at most once, at least once, exactly once) and supports persistent sessions that shorten reconnection after a dropout. Local routes keep working when the internet does not, which matters on a farm.
Route 3: file exports and scheduled reports
Some systems can only export a CSV file or email a report. That is fine for data that changes slowly, such as weights, yields and machine hours. A daily import brings those numbers into the same farm monitoring system as your live readings, without pretending they are real time.
Route 4: dry contacts when there is no way out
When a device has no data connection at all, look for an alarm relay, a volt-free contact or a status light. A cheap input module can turn “alarm relay closed” into a message, and a bolt-on sensor can measure what the device will not share: a temperature probe on the milk tank, or a current clamp on a fan motor. This route is how older kit joins a modern farm monitoring system without being replaced.
| Route | Best for | Typical delay | Works without internet? | Main risk |
|---|---|---|---|---|
| Vendor cloud API | Weather stations, herd software, telematics | Minutes | No | Terms, fees or API changes |
| Modbus or MQTT on site | Store, ventilation and pump controllers | Seconds | Yes, on site | Needs wiring and a gateway |
| File export or report | Weights, yields, machine hours | Hours to a day | Partly | Manual steps creep back in |
| Dry contact or bolt-on sensor | Older kit with alarms only | Seconds | Yes, on site | Only on/off or a single reading |
A Farm Monitoring System Architecture That Suits Most Farms
A reliable farm monitoring system has five layers, and on a small farm several of them can live on one box. Thinking in layers lets you swap one part later without rebuilding everything.
Layer 1: sensors and the devices you already own
This is your audit list plus a few new sensors to fill genuine gaps. Keep it boring: proven devices, replaceable batteries, and a clear label on each sensor saying what it measures and where.
Layer 2: a gateway in the yard
A gateway is a small, always-on computer, such as an industrial mini PC or a single-board computer in a proper enclosure. It collects readings from local devices, buffers them when the internet drops, and forwards them. Put it somewhere dry with power protection. Crucially, run the most important alarms on the gateway itself, so a broadband fault does not silence a grain store alert. Our IoT solutions team treats this box as the heart of a farm monitoring system, not an accessory.
Layer 3: a message broker
An MQTT broker, with Mosquitto a common open-source choice, works like a post room. Every device and integration publishes readings to named topics such as store2/bay3/temp, and anything that needs them subscribes. The broker separates sources from consumers, so adding a new dashboard or alert does not mean touching the sensors.
Layer 4: storage for readings
Readings are time series: a value, a time and a label. A time-series database such as InfluxDB or TimescaleDB stores them efficiently and answers questions like “average store temperature per day since harvest”. A single farm with a few dozen sensors produces modest volumes, so keep raw readings for months and daily summaries for years.
Layer 5: dashboards and alerts
This is the part people see. It can be Home Assistant, Grafana, a vendor app or a mix. The key design decision is where alerts are raised. Keep critical alerts as close to the sensor as possible, and duplicate them in the cloud.
| Layer | Job | Common options | Where it runs |
|---|---|---|---|
| Sensors and devices | Measure | Existing kit plus LoRaWAN or wired sensors | Field, store, shed |
| Gateway | Collect, buffer, local alarms | Industrial mini PC, single-board computer | Farm office or yard |
| Message broker | Route readings | MQTT broker such as Mosquitto | Gateway |
| Storage | Keep history | InfluxDB, TimescaleDB, vendor cloud | Gateway or cloud |
| Dashboards and alerts | Show and notify | Home Assistant, Grafana, Node-RED, vendor apps | Gateway, cloud, phone |
Connectivity: Getting Farm Monitoring System Readings Back From the Field
Connectivity shapes more farm monitoring system designs than any sensor choice. A sensor that cannot report is worse than no sensor, because it gives false comfort.
What national coverage figures hide
Ofcom’s Connected Nations 2025 report puts good outdoor 4G coverage from at least one operator at around 96% of the UK landmass. “At least one” does a lot of work in that sentence. Ofcom’s figures for England show indoor 4G coverage in rural areas of 77% to 85% depending on the operator, against 97% to 99% in urban areas. Stores, sheds and parlours are exactly where indoor coverage matters, so test on site with the SIM you plan to use.
LoRaWAN for battery sensors across the farm
LoRaWAN is a low-power, wide-area radio network for battery devices. The LoRa Alliance quotes a range of up to 15 km in rural areas, which suits soil probes, trough levels, gate sensors and store probes that send small readings infrequently. One gateway on a tall building can cover much of a farm.
LoRaWAN is not suited to images or large files, and the airtime rules matter. On The Things Network’s free community sandbox, a fair use policy limits each device to 30 seconds of uplink airtime and 10 downlink messages per day. A private network is not bound by that policy but must still stay within the regulated duty-cycle limits, which in most regions are 1% on the join frequencies.
The arithmetic below shows why the reporting interval matters on the community network. It simply divides the 30-second daily allowance by the number of messages a sensor sends each day.
A sensor at the edge of range transmits more slowly, and so uses far more airtime per message, than one near the gateway. A probe that reports every five minutes from the far side of the farm can therefore exhaust its allowance. For most farm readings, every 15 to 30 minutes is plenty, and a well-designed sensor sends an extra message the moment a threshold is crossed.
4G routers, Wi-Fi bridges and satellite
For the gateway’s own internet link, a 4G router with an external antenna mounted high often beats the farmhouse broadband for resilience. Point-to-point Wi-Fi bridges can carry a connection from the house to a distant yard. Where neither works, satellite broadband is now a practical backhaul, and our satellite-as-a-service rural failover checks cover what to test. For larger sites with several yards, a proper network design saves a lot of trial and error.
Store and forward when the link drops
Design every link to fail. The gateway should hold readings locally while the internet is down and send them when it returns, and local alarms should still reach someone through a backup route such as an SMS modem. A farm monitoring system that goes silent during a storm has failed at the one moment it was needed.
Choosing Software for a Farm Monitoring System
Software is where many farm projects stall, because there are too many options. The answer is usually a small combination of tools rather than one product.
Keep your existing farm software as the system of record
Whatever you use for field records, herd records and compliance stays in charge of those records. The farm monitoring system reads from it where it can, for example which bay holds which crop, but does not replace it.
Home Assistant for small, mixed setups
Home Assistant is open-source home automation software that puts local control and privacy first, with more than 1,500 integrations. Many of those integrations cover exactly the consumer and prosumer kit found on farms: weather stations, smart plugs, cameras and battery sensors. It runs on a small box, works without the internet, and sends phone notifications. It suits a farm with a modest number of devices and one person who enjoys tinkering.
Node-RED for glue logic
Node-RED is a low-code tool for event-driven applications. You wire inputs, logic and outputs together as a flow in a web browser, and its community has built more than 5,000 nodes and flows, including Modbus, MQTT and many vendor APIs. It is the natural home for rules such as “if the tank temperature is above threshold and no wash cycle is running, alert the herd manager”.
Grafana and a time-series database for trends and alerting
Grafana draws dashboards from time-series databases and can raise alerts itself. Its contact points include email, Telegram, Microsoft Teams, Slack, webhooks and MQTT, and its alert rules have a No Data state that can fire when a sensor stops reporting. Pair it with InfluxDB or TimescaleDB for history. Our data visualisation services team builds farm monitoring system dashboards on this pattern.
Vendor platforms: leave them in charge where they are good
A robotic milking system or a store controller may already provide excellent alerts. Do not rebuild what works. Copy the key readings into your farm monitoring system for a combined view and a long history, and leave the vendor’s own alarms switched on as a second line.
| Software | Best at | Skills needed | Licence | Watch out for |
|---|---|---|---|---|
| Home Assistant | Mixed consumer kit, phone alerts | Low to medium | Open source | Farm-wide LoRaWAN needs extra set-up |
| Node-RED | Rules and integrations | Medium | Open source | Flows need documenting |
| Grafana plus a time-series database | History, trends, alert rules | Medium | Open source, with paid cloud tiers | Needs a data source to be useful |
| Vendor apps and portals | Their own devices | Low | Bundled or subscription | Data locked to one brand |
| Spreadsheet | Weekly manual readings | Low | Already owned | No alerts, no live data |
Farm Monitoring System Alerts People Will Not Ignore
Alerts are where a farm monitoring system succeeds or fails. Too few, and losses go unnoticed. Too many, and people mute the phone.
Set thresholds from published targets, not guesses
Use published guidance wherever it exists. AHDB’s grain cooling targets are clear: cool grain to below 15°C within two weeks of harvest, below 12°C within three to four months, and below 5°C by the end of December. Its moisture targets note that food safety risk rises above 14.5% moisture content in cereals, and that a localised moisture rise of 2% or more in a week needs investigating.
For milk, the retained hygiene rules in Regulation (EC) No 853/2004 require milk to be cooled immediately to not more than 8°C where collection is daily, or not more than 6°C where it is not. Those numbers make obvious alert thresholds for a farm monitoring system.
| What to watch | Alert when | Source of the number | Who gets it |
|---|---|---|---|
| Grain temperature after harvest | Still 15°C or above at day 14 | AHDB cooling targets | Store manager |
| Grain temperature in winter | Still 5°C or above at the end of December | AHDB cooling targets | Store manager |
| Grain moisture | A local rise of 2% or more in a week | AHDB moisture targets | Store manager |
| Bulk milk temperature | Above 8°C (daily collection) or 6°C (not daily) | Regulation (EC) No 853/2004, as retained | Herd manager, then owner |
| Vaccine fridge | Outside 2°C to 8°C | FETF 2026 vaccine fridge specification | Stockperson |
| Any critical sensor | No reading for two hours | Your own setting, using a No Data rule | Whoever maintains the farm monitoring system |
| Water header tank | Level falling faster than it refills | Your own water meter readings | Stockperson |
Why cooling matters: the humidity inside the heap
AHDB explains that stored grain exchanges water with the air around it until it reaches an equilibrium relative humidity (ERH), and that mould growth and mite reproduction stop below 65% ERH. Its figures for wheat at 14.5% moisture content show how much temperature alone moves that number.
The same grain sits on the wrong side of the 65% line at 25°C and comfortably below it at 5°C. That is why a temperature trend from the store is worth more than any single spot reading, and why grain is the first job for many arable farm monitoring system builds.
Alert on silence, not just on values
The most dangerous failure is a sensor that stops reporting while the dashboard keeps showing its last good value. Set a “no data” rule on every critical sensor. Grafana’s No Data state exists for exactly this, and a Node-RED flow can do the same with a simple timer.
Use rate of change for slow problems
Heating grain and a leaking tank are slow problems that a fixed threshold catches late. AHDB notes that rising store temperatures can indicate fungi, sprouting or grain weevils. A rule that compares today’s average with last week’s catches a slow climb days before it crosses a hard limit.
Water works the same way. Suppose, as an illustration, a shed’s header tank holds 2,000 litres and your meter shows the shed drawing around 250 litres an hour in hot weather. If the supply fails, the tank lasts 2,000 ÷ 250 = 8 hours, so an alert on “level falling for 30 minutes” buys most of a working day.
Escalate, and respect quiet hours
Send the first alert to the person on duty, escalate to a second person if nobody acknowledges within, say, 15 minutes, and reserve night-time calls for alarms that cannot wait. A store temperature creeping up by half a degree is a morning job. A milk tank that has stopped cooling is not.
How to Build a Farm Monitoring System in Five Stages
A farm monitoring system is best built in stages that each deliver something useful on their own, rather than as one large project with a big reveal.
Stage 1: pick one problem worth solving
Choose the single loss that worries you most, such as grain heating, milk cooling or water supply, and build for that first. A working alert on one real problem teaches more than a dashboard of twenty readings.
Stage 2: connect what you already own
Use the audit to connect the devices relevant to that problem. Put the gateway and broker in place now, even if only two devices use them at first.
Stage 3: fill one gap with one new sensor
Add a new sensor only where the audit shows a genuine gap, such as a level sensor in a header tank or a probe at the right depth in a store. Adding one device at a time keeps faults easy to trace.
Stage 4: alerts first, dashboards second
Get alerts working and tested before building dashboards. Test by forcing each alarm: warm a probe in your hand, lift a float switch, unplug a sensor to prove the no-data rule fires.
Stage 5: review after a season
After a harvest or a winter, review which alerts fired, which were ignored, and which losses the farm monitoring system missed. Adjust thresholds, remove noisy alerts, then pick the next problem.
| Stage | Planning estimate | Output you can check |
|---|---|---|
| 1. Pick one problem | A morning | One named loss and one owner |
| 2. Connect existing kit | One to two weeks | Readings arriving at the gateway |
| 3. Fill one gap | About a week per sensor | New sensor reporting reliably for seven days |
| 4. Alerts, then dashboards | One to two weeks | Every alarm tested by forcing it |
| 5. Review | After a season | Alerts trimmed, next problem chosen |
The durations are planning estimates for a part-time build alongside farm work, not measurements. Harvest and lambing will stretch them, which is another reason to start a farm monitoring system build in a quiet month.
Worked Example: A Farm Monitoring System for a Mixed Farm
To make the costs concrete, here is an illustrative mixed arable and dairy farm. The farm, its kit and every price below are planning assumptions for September 2026, not quotes.
What the farm already owns
The farm has an on-floor grain store with temperature cables and a controller that has a Modbus port, a bulk milk tank whose controller has an alarm relay but no data output, a Davis weather station already uploading to WeatherLink, two water header tanks with float valves and no sensors, herd software with a daily export, and CCTV in the yard.
What the farm monitoring system does with it
The gateway polls the store controller over Modbus and raises alerts against the AHDB targets. The tank controller’s alarm relay is wired to an input module, and one LoRaWAN temperature probe on the tank gives a continuous milk temperature reading with an 8°C alert.
Weather readings arrive from the WeatherLink API every 15 minutes and drive a frost alert. Two LoRaWAN level sensors watch the header tanks for a falling level. Herd weights and yields arrive once a day from the export, with no alerts. The CCTV recorder sends a night-time motion event from the yard gate to the duty phone.
| Item (planning assumption) | Quantity | Unit cost | Total |
|---|---|---|---|
| Industrial gateway (mini PC) | 1 | £450 | £450 |
| Outdoor LoRaWAN gateway | 1 | £650 | £650 |
| 4G router with external antenna | 1 | £350 | £350 |
| UPS for gateway and router | 1 | £150 | £150 |
| LoRaWAN tank level sensor | 2 | £220 | £440 |
| LoRaWAN temperature probe for the milk tank | 1 | £120 | £120 |
| Relay input module for the tank alarm | 1 | £90 | £90 |
| New hardware subtotal | £2,250 | ||
| Integration and set-up, five days at £600 | 5 | £600 | £3,000 |
| 4G data SIM, first year at £10 a month | 12 | £10 | £120 |
| Year one total | £5,370 |
The software stack in this example is open source (Mosquitto, Node-RED, InfluxDB and Grafana, or Home Assistant for a simpler build), so there is no licence line. Check separately whether your weather station’s data plan includes API access. From year two, assume £120 a year for the SIM plus one maintenance day at £600, or £720 a year. Over three years that is £5,370 + (£720 × 2) = £6,810, or about £189 a month.
Two things stand out. First, labour is the largest line, which is why a farmer who enjoys this work can cut the cash cost to £2,250 of hardware plus £360 of data over three years, or £2,610, in exchange for their own time. Second, the store controller the farm already owned did most of the grain work. For comparison, the FETF 2026 expected average cost of a new crop storage sensor system on its own was £4,818.
What a Farm Monitoring System Costs, and What FETF 2026 Paid For
Published grant lists are a useful price reference, because they state an expected average cost for each item. They also show what Defra expects monitoring kit to do.
Reference prices from the FETF 2026 item lists
The Farming Equipment and Technology Fund 2026 opened on 17 March 2026 and closed at midday on 12 May 2026. It offered grants of between £1,000 and £25,000 for each of three themes (productivity, slurry, and animal health and welfare) to farm businesses in England. The monitoring items in its productivity list and animal health and welfare list are summarised below.
| FETF 2026 item | Expected average cost | Grant | What the specification asks for |
|---|---|---|---|
| FETF315SH automatic health and welfare remote monitoring (sheep) | £17,549 | £8,774 | Record, transmit and process animal data and trigger alerts; cost based on one base unit and around 200 tags, collars, bands or boluses |
| FETF117 crop storage sensor system | £4,818 | £1,927 | Control unit sends information to an existing phone, tablet or computer, with alerts on temperature or humidity |
| FETF253A automatic monitoring main unit, pig housing | £2,301 | £1,150 | Integrate ammonia, temperature, humidity, water disappearance and feed bin weight, with alarms |
| FETF118 digital weather station | £1,804 | £721 | Link wirelessly to a computer; software licence and calibration fees not covered |
| FETF271A precision water meter for poultry chicks | £943 | £471 | Transmit flow rate data to an alarm system |
| FETF213 remote soil moisture sensing | £796 | £318 | Bluetooth or similar link to a base station or phone; software not eligible |
| FETF277SH temperature and humidity sensor, livestock housing | £164 | £82 | Integrate with a new or existing climate or housing management system |
Notice how often the specifications talk about integration with what a farm already has: an existing phone, tablet or computer, or a new or existing management system. That is the same principle this guide is built on. A new sensor should report into your farm monitoring system, not start another island.
Software and subscription costs
Open-source tools such as Home Assistant, Node-RED, Grafana, Mosquitto and InfluxDB’s open-source edition carry no licence fee, so the costs are hardware, time and support. Vendor platforms usually charge per device or per site. FETF 2026 generally excluded ongoing subscriptions, and several livestock items that depend on subscription software required a minimum five-year subscription paid in advance. That is a useful reminder to price the whole life of any platform, not just the first year.
Funding after FETF 2026
FETF 2026 is closed. Defra’s funding for farmers, growers and land managers page lists current schemes. If you plan to apply to a future round, check the timing rules before buying: the FETF 2026 application guidance said applicants must not buy any item in their application before receiving a grant funding agreement, unless they paid a refundable deposit.
Security and Data Ownership in a Farm Monitoring System
A connected farm has more doors than a paper one. Security for a farm monitoring system is mostly about a few habits, applied to every device.
Change every default password
The UK’s consumer connectable product security regime, in force since 29 April 2024 under the Product Security and Telecommunications Infrastructure Act 2022, bans universal default and easily guessable passwords. It also requires manufacturers to publish how to report security issues and the minimum period for security updates. The regime covers consumer connectable products, so professional farm equipment sold to businesses may fall outside it, and older kit predates it entirely. Check every device yourself: change default passwords, keep them in a password manager, and note each vendor’s update period.
Never open ports to the internet
Remote access is the biggest risk in a home-built farm monitoring system. Do not forward ports on the router to reach a controller or a dashboard from outside. Use a VPN or the vendor’s secure remote access instead, and keep monitoring devices on their own network segment, away from the farm office PCs. Our cybersecurity team can review a farm monitoring system like this before it goes live.
Know who owns the readings
Read the data terms for every vendor cloud you connect. Check who can see your readings, whether they can be shared or sold, whether you can export all of them, and what happens if you stop paying. Your farm monitoring system should keep its own copy of the readings that matter, so you are never locked in.
Plan for power cuts and backups
Put the gateway, the router and the LoRaWAN gateway on a UPS, and back up configuration and data off site. A power cut is exactly when fans, pumps and cooling stop, so the farm monitoring system needs to survive long enough to send the alert.
Common Farm Monitoring System Mistakes
Most failed farm monitoring system projects fail for the same handful of reasons, and none of them is a lack of technology.
Buying a platform before listing what you own
A new all-in-one platform can duplicate sensors you already have and leave your best data, such as the store controller, outside it. Do the audit first.
Too many alerts, too early
Twenty alerts in the first week teaches everyone to ignore the phone. Start with the few that matter, test them, and add more only when those are trusted.
Sensors in the wrong place
A store probe near the fan duct reads cool while the far corner heats. Follow the AHDB guidance on measuring where cooling takes longest, and record at the same locations each time.
No plan for batteries and updates
Battery sensors die, usually in winter. Keep a list of every battery device, its expected battery life and its last change, and let the no-data rule tell you which one has gone quiet.
Nobody owns the farm monitoring system
A farm monitoring system needs a named owner who fixes it when it breaks, even if the build was done by a contractor. Without one, the first fault becomes the last.
When to Bring In Help
There are three sensible ways to build the farm monitoring system described here, and many farms mix them.
Do it yourself
If someone on the farm is comfortable with small computers and wiring, the whole farm monitoring system in this guide can be built in-house. Budget time for learning, and document everything so the knowledge does not leave with one person.
Use an integrator
An integrator can survey the farm, connect the awkward controllers, set up remote access safely and hand over a documented system. This suits farms with several sites, older industrial controllers or a requirement for an audit trail. Our earlier piece on agritech infrastructure describes the network and cloud side of that work.
Let the vendor do it
For a single problem that one vendor already solves well, such as a store controller with good alarms, the simplest answer may be to switch on the features you already pay for. You can still pull its readings into a wider farm monitoring system later.
Farm Monitoring System FAQs
Can I build a farm monitoring system without Wi-Fi across the farm?
Yes. Battery sensors can report over LoRaWAN, which the LoRa Alliance says can reach up to 15 km in rural areas, to one gateway on a tall building. Only that gateway needs an internet link, which can be 4G, a Wi-Fi bridge to the farmhouse or satellite.
Do I have to replace my existing sensors?
Usually not. Most devices can join through a vendor API, a local protocol such as Modbus, a file export or an alarm relay. Replace a device only when none of those routes works and the reading genuinely matters.
How much does a farm monitoring system cost?
It depends on what you already own. In the worked example above, connecting existing kit and adding four sensors cost £5,370 in year one and £6,810 over three years with an integrator, or £2,610 in cash if the farm does the work itself. The FETF 2026 lists give reference prices for new kit, from £164 for a housing temperature and humidity sensor to £4,818 for a crop storage sensor system.
Which software is best for a farm monitoring system?
There is no single best tool. Home Assistant suits small mixed setups, Node-RED handles rules and integrations, and Grafana with a time-series database handles history and alerting. Keep your existing farm management software as the record of what happened in the field.
How often should sensors report?
For most farm readings, every 15 to 30 minutes is enough, with an immediate extra message when a threshold is crossed. On The Things Network’s community sandbox, the 30-second daily airtime allowance makes very frequent reporting impractical for distant sensors.
Is a farm monitoring system worth it on a small farm?
If one prevented loss, such as a batch of heated grain or a tank of warm milk, would cover the cost, then yes. Start with that one problem and the kit you own, and the spend stays small.
References and Further Reading
Farming Equipment and Technology Fund (FETF) 2026 (GOV.UK)
FETF 2026 productivity items and specifications (GOV.UK)
FETF 2026 animal health and welfare items and specifications (GOV.UK)
Apply for the Farming Equipment and Technology Fund 2026 (GOV.UK)
Funding for farmers, growers and land managers (GOV.UK)
Regulations: consumer connectable product security (GOV.UK)
Cooling targets for grain storage (AHDB)
Grain storage moisture targets for cereals and oilseed rape (AHDB)
Regulation (EC) No 853/2004, Annex III (legislation.gov.uk)
Connected Nations 2025 (Ofcom)
What is LoRaWAN (LoRa Alliance)
Duty cycle and fair use policy (The Things Network)
MQTT: The Standard for IoT Messaging
Node-RED: low-code programming for event-driven applications
Contact points (Grafana documentation)
