Soil and plant monitoring on the cheap: a soil moisture monitoring system you own

If you want a soil moisture monitoring system that won't cost a fortune or die when an app gets pulled, build it on cheap capacitive Zigbee sensors and Home Assistant: battery-powered, wireless, and all available in Australia. Sensors run about $20 to $35 each, the coordinator is a one-off dongle, the software's free, and there's no monthly fee and no vendor bridge to lock you in. That's the verdict. Here's why the branded, app-locked soil gadget is usually the wrong buy, and exactly how I'd put the cheap version together.

The one buying decision that matters: capacitive, not resistive

Before anything else, bands, hubs, brands, get this right or the rest is wasted. There are two kinds of cheap soil probe, and one of them is rubbish.

Resistive probes are the bare two-pronged forks you see for a couple of dollars. They measure moisture by passing current between two exposed metal prongs sitting in wet, salty, often fertilised soil. That is a corrosion cell. Within a season the prongs oxidise, the readings drift, and you're chasing a fault that's really just a dissolving sensor. Don't buy them. I've watched people blame their whole setup when the only problem was the cheapest part rotting in the dirt.

Capacitive probes have no exposed metal; the sensing element is sealed behind the probe body and reads moisture as a change in capacitance. Nothing corrodes, so they last for years in the ground. They cost a few dollars more and they're the only ones worth burying. Every sensor I'd actually deploy is capacitive. If a listing doesn't say "capacitive," assume it's the corroding kind and move on.

Why Zigbee, and why that beats a branded kit

Here's the trap with the tidy boxed soil-sensor kits: most of them only talk to their own bridge and their own app, and that app wants an account, sometimes a subscription, and an internet connection to tell you whether your tomatoes are dry. When the brand discontinues the line or shuts the cloud down, and consumer IoT brands do this constantly, your sensors are landfill. You rented a garden gadget; you didn't own a system.

Zigbee flips that. It's an open, low-power wireless standard, and a standards-based Zigbee sensor pairs to any Zigbee coordinator. So one $30 to $50 USB dongle plugged into Home Assistant becomes the single brain for every Zigbee sensor in the garden, whatever brand they are. One coordinator, many sensors, no per-device app, no cloud dependency. That's the same "own it, don't rent it" logic behind running your own hub instead of leaning on someone else's servers, and with a garden the difference is whether your watering logic still runs when the NBN drops out at dinnertime.

Zigbee is also genuinely frugal on power, which is the other half of "cheap." The sensors run for a year or more on a small coin or AAA cell because the radio sleeps almost all the time and wakes briefly to send a reading. Soil moisture moves slowly: a reading an hour tells you everything a reading a minute would, at a fraction of the battery. Anything claiming constant live readings off a tiny battery is either lying or about to flatten it.

What the whole system actually costs

No mystery, no "contact us for pricing." Roughly, in Australian dollars:

So a real four-zone garden monitor lands somewhere around $150 to $200 all in, once, with no ongoing cost. A branded equivalent with four cloud-tied sensors gets there on hardware alone and then keeps charging you, and it dies on the brand's timetable, not yours.

How I'd put it together

Work down this list. It's a weekend job, not a project.

  1. Stand up Home Assistant. Flash it to a Pi or run it on a spare mini PC. It's free and it's the hub everything else hangs off.
  2. Plug in the coordinator and add ZHA or Zigbee2MQTT. Both are free integrations that turn the dongle into your Zigbee network. ZHA is the simpler start; Zigbee2MQTT gives you broader device support. Either is fine for soil sensors.
  3. Pair the sensors near the hub first. Put each one into pairing mode close to the coordinator, confirm it joins, label it for its zone, then take it out to the garden. Pairing up close and deploying after saves a lot of "why won't it join from 30 metres away."
  4. Push the probe in properly. Sensing element fully in the root zone, not just the dry crust on top, and away from the very edge of a pot where it reads drier than the plant feels. Where you put it decides whether the number means anything.
  5. Set thresholds and alerts. Once you've watched a zone for a week you'll know its normal wet-to-dry swing. Set a low threshold below that and have Home Assistant send a phone notification when a zone crosses it. Now you find out the natives are dry from your phone, not from a dead plant.

The two things that trip people up

Range. Zigbee is short-range by design and it does not love wet soil, garden beds and brick between the sensor and the coordinator. If a far corner won't connect, don't fight it: drop a mains-powered Zigbee device (a plug in the shed, a globe on the back verandah) between the two. Every mains Zigbee device is also a repeater, so it builds a mesh and the garden sensors hop through it. Adding repeaters is how you extend Zigbee outward; it's cheaper and more reliable than buying a "stronger" sensor.

Scale. Zigbee is the right tool for a house block and a garden. It is the wrong tool the moment you're talking paddocks, a back dam, or anything past the far fence. Zigbee simply won't span that, and stacking repeaters across a property is a losing game. That's a different class of problem solved by long-range radio sensors: a proper farm and property IoT monitoring setup. Same idea, cheap sensors and real alerts, just a radio that's made to cross distance. If you're watching water rather than soil, the same family covers a stock trough or dam level, and even tank monitoring where there's no mobile signal at all. Use Zigbee where Zigbee fits; reach for the bigger radio where it doesn't.

The payoff: it waters the garden for you

Monitoring on its own is useful: you stop guessing and you stop overwatering. But because you own the hub, you get the part the branded kits charge extra for or can't do at all: automation. In Home Assistant it's a couple of lines of logic: if this zone drops below its threshold and there's no rain forecast, open a Zigbee or wifi valve for ten minutes. It runs locally on your own box, so it keeps working through an internet outage, which a cloud-only system cannot promise. Set it once, set it right, and the garden looks after itself.

Soil moisture monitoring system: common questions

What is the cheapest soil moisture monitoring system that actually works?
A handful of capacitive Zigbee soil sensors paired to Home Assistant through a single Zigbee coordinator. The sensors run about 20 to 35 dollars each in Australia, the coordinator is a one-off 30 to 50 dollar dongle, and Home Assistant is free on a Raspberry Pi you may already own. There is no monthly fee and no per-sensor app account. Crucially, buy capacitive sensors, not the cheap two-pronged resistive probes. Resistive ones corrode in the soil within a season and read rubbish.
Do I need a vendor's hub or bridge for soil moisture sensors?
No, and that is the whole point of going Zigbee. A standards-based Zigbee sensor pairs to any Zigbee coordinator, so one 30 to 50 dollar dongle on Home Assistant talks to every Zigbee sensor in the garden regardless of brand. Vendor-locked sensors that only work with their own bridge and cloud account are the trap: when the brand drops the product or the app, your sensors become landfill. Own the coordinator, not a subscription.
How long do battery-powered soil sensors last?
A good capacitive Zigbee sensor on a CR2 or AAA cell reports every hour or so and lasts roughly a year to eighteen months. Soil moisture changes slowly, so you do not need second-by-second readings; a reading an hour is plenty and is what makes the battery last. If a sensor is flattening a battery in weeks, it is reporting far too often or it is a cheap resistive probe holding the radio open.
Will Zigbee soil sensors reach the back of the yard?
Often, but Zigbee is short-range and does not love wet soil or brick walls between the sensor and the coordinator. The fix is mesh: any mains-powered Zigbee device, such as a smart plug or globe, acts as a repeater and extends the network outward. Drop one in the shed or on the back verandah and the garden sensors hop through it. For a paddock or a block beyond Zigbee range entirely, that is a job for a long-range radio sensor, not Zigbee.
Can a soil moisture monitoring system control my irrigation automatically?
Yes, and this is where owning the stack pays off. In Home Assistant you write a simple automation: if a zone's soil moisture drops below your threshold and no rain is forecast, switch on a Zigbee or wifi valve for a set time. Because everything lives on your own hub, the logic runs locally and keeps working even if your internet drops, which a cloud-only branded system cannot promise.
Are these soil sensors available in Australia?
Yes. Capacitive Zigbee soil sensors, Zigbee coordinator dongles and Home Assistant gear are all readily available from Australian electronics retailers and the usual marketplaces, and they run on the same Zigbee standard sold worldwide. Stick to capacitive sensors and a well-supported coordinator chipset, and you can build the whole system without importing anything exotic.

The bottom line

A soil moisture monitoring system doesn't need to be expensive or rented. Capacitive sensors so they don't rot, Zigbee so they're not chained to one brand, Home Assistant so the brains and the data are yours, and battery plus wireless so there's nothing to trench in. Build it that way and you've got something you control, that costs almost nothing to run, and that you can grow from three pots to the whole yard without re-buying anything.

Want it built and tuned properly the first time, with sensors placed where the readings mean something, range sorted, watering automated, and no app account or lock-in? That's the kind of job we do, and we'll happily point you at the cheap DIY path if that suits you better. Tell us what you're growing and we'll map it out.