Catch water leaks in hours, not weeks
A leak on a shared well or in an apartment block can run for weeks before anyone notices — and the first sign is usually an unexpectedly high bill. As the installer, you are the one who gets the call. Datakubo lets you catch those leaks early, whichever meter you install — and with nothing to configure.
It works in two layers. Some meters do the detection themselves, in firmware: the alarm arrives pre-triggered on the uplink and Datakubo surfaces it as-is. The meters that only count litres are covered too — for those, Datakubo computes the alarms from the consumption data itself. There are no rules to define and no thresholds to tune in either layer, and your client sees the same plain-language alert under your brand.

Layer 1 — Alarms the meter already raises: nothing to configure
Some meters do the detection themselves, inside the firmware. If yours is one of them there is nothing to wire up: the alarm arrives pre-triggered on the same uplink and Datakubo surfaces it as-is, in real time. And if your meter raises none of these? Keep reading — that is what layer 2 is for.
Leak / continuous flow
The moment the meter senses water flowing without stopping — the classic signature of a burst pipe or a running cistern — it raises its own leak alarm. There is no threshold to set: the alarm arrives pre-triggered straight from the device firmware.
Burst
Compatible meters detect an internal flow spike consistent with a burst pipe and report a burst alarm on the same uplink. Datakubo just surfaces it in the portal — the detection itself already happens in hardware.
Reverse flow
If water starts flowing backwards — a common sign of a bad repair or a misinstalled valve — the meter flags a reverse-flow alarm automatically, with no setup on your side.
Overflow
When flow exceeds the range the meter can measure accurately, the device raises an overflow alarm on its own. It is a factory-set signal, not a rule you need to tune.
Wrong installation
The meter checks its own orientation and installation and flags it when something is mounted incorrectly — useful for catching installation mistakes before they turn into a support call.
Low battery
The meter reports its own battery state on every reading. Once it drops below a safe level, the low-battery alarm fires on its own — no threshold for you to define.
Layer 2 — Your meter raises no alarms? Datakubo raises them
Plenty of LoRaWAN meters only count litres: no alarms from the factory. With those, the detection does not disappear — it moves. Datakubo computes it from the consumption itself, with the same zero clicks of setup. Tell us the model you install and we will tell you exactly which alarms the meter gives you and which ones the platform adds, before you pay for anything.
Zone leak, with no alarm from the meter
Datakubo compares the main meter against the sum of the individual meters it feeds. If more than 10% of the flow, or more than 0.1 m³, goes missing — and the gap repeats across three consecutive checks — a leak alarm opens with its own severity level. No meter needs to know anything: a cumulative total is enough.
Consumption spike
Every reading is compared against the last 30 days of that same meter. A jump well above its own pattern is flagged on its own, without you defining any threshold: the threshold is that meter’s own history.
Sustained elevated use
A meter running above its own normal for days on end gets flagged even when no single reading looks alarming. That is the signature of the slow leak — the one that eats a whole bill before anyone sees it.
Main meter stopped
If the community main meter reports zero for many consecutive readings, something is wrong: a jammed meter, a dead pump, or a frozen pipe. The platform raises it without anyone watching the screen.
Offline / no data
Meters that stop reporting altogether — flat battery, network gap, or tampering — are flagged automatically, whatever the brand.
Hours, not weeks — in both layers
Hardware alarms arrive on the same uplink; the computed ones are evaluated twice a day. Either way we are talking hours rather than weeks, and in neither case is there anything for you to configure.
Non-revenue water (NRW): how much is going missing, and where
Non-revenue water is the water that enters a network and never reaches any customer meter — the gap between the main meter and the sum of the individual ones. It is the number water utilities and community managers are actually judged on, and it is the number a leak eventually shows up in. Datakubo calculates it for you every day, in m³ and as a share of the water that entered the zone, with no spreadsheet and no meter reading round.
- The analytics screen shows the period’s NRW in m³ and as a percentage of intake, with the gap between the main meter and the sum of the sub-meters drawn day by day.
- NRW is not the same as a leak. Datakubo splits the gap into three parts: habitual unmetered use (irrigation, common areas), meters that failed to report, and the genuine unexplained loss.
- Alarms are raised on the unexplained part, not on the total — so an irrigation line that has always been there does not page anyone every morning.
- Every zone carries its own NRW figure, so you learn which branch is losing water instead of only that the site as a whole is.
What your client actually sees
Alerts are written for the person reading them, not for an engineer. A community manager sees something like: "Meter 14 (Building B) has reported a leak alarm. This often means a burst pipe or a fixture that will not stop running. We recommend checking the property." No raw payloads, no thresholds, no jargon — just a plain-language heads-up under your brand, with the meter, the location and a suggested next step.
How AquaLinks uses leak alerts
AquaLinks, an independent LoRaWAN installer in Spain, runs Datakubo across three private shared-well communities and more than 120 meters — starting with a first community of 66 meters, followed by a second of 39. In that first community, a B Meters device raised a leak alarm on its own, and Datakubo surfaced it in the portal the moment it arrived — residents get automatic leak alerts and see their own consumption from their phone. The installation won Best Product at ExpoEbre 2025.
How to set it up
Connect your meters
Point a TTN or ChirpStack webhook at Datakubo. Each LoRaWAN meter auto-registers on its first uplink — no manual entry per device.
That is it — alarms surface automatically
There are no thresholds to set. If the meter firmware raises leak, burst, reverse-flow, overflow, wrong-installation or low-battery alarms, Datakubo picks them up the moment they arrive on the next uplink. If it raises none, the platform computes the alarms from consumption instead — nothing for you to touch there either.
Let your clients see the alerts
Invite the community manager and residents to your branded portal. From then on, alerts surface automatically — and each resident sees only their own meter.
Works with the meters you already install
- B Meters (GSD8-RFM, GMDX-RFM, and other LoRaWAN models)
- Milesight water meters
- Dragino LoRaWAN water sensors
- Any pulse-counter LoRaWAN meter (generic payload support)
- Meters with no factory alarms — Datakubo computes them from consumption
Frequently asked questions about non-revenue water
What is non-revenue water (NRW)?
Non-revenue water is the difference between the water that enters a network — measured by the main meter — and the water recorded by the customer meters it feeds. It is expressed in m³ and as a percentage of intake. In Spanish it is called agua no registrada (ANR). It is water that was supplied but that nobody was billed for.
How does Datakubo calculate non-revenue water?
NRW = main meter consumption − the sum of the end-point meters under it, for the same period and time-aligned so a late-arriving reading cannot fake a gap. Datakubo evaluates it twice a day for every zone and shows the result in m³ and as a share of the water that entered the zone. Nothing to configure: it comes from the meter hierarchy you already set up.
Does high non-revenue water mean there is a leak?
Not on its own. A gap can be a genuine leak, but it can equally be legitimate unmetered use — irrigation, common areas, cleaning — or meters that simply failed to report that day. Datakubo separates those three before raising anything, subtracting the zone’s habitual unmetered use and estimating what the offline meters would have counted, so the alarm lands on the part that is genuinely unexplained.
What is an acceptable NRW percentage?
There is no single number: a community with shared irrigation and common areas legitimately runs a higher share than a block of flats where every drop passes a meter. What matters is your own baseline and its trend. Datakubo learns each zone’s habitual level from the last 14 days and flags the departure from it, rather than comparing you to a figure from someone else’s network.
Do I need special meters to measure non-revenue water?
No. Any LoRaWAN meter that reports a cumulative total works, whatever the brand and whether or not it ships its own alarms. What NRW needs is a hierarchy — one main meter and the individual meters under it — which you define once in the platform.