Retrofit sensors
Clamp-on and inline units fit the mechanical meters and junctions already in the ground. No wholesale meter replacement programme, no waiting for a capital cycle.
Water network telemetry
Nabaflow retrofits pressure and flow sensors onto the meters and junctions a utility already owns. Continuous telemetry replaces quarterly reading rounds, and anomaly models narrow a suspected loss to a pipe segment — before it reaches the surface.
Retrofit hardware · NB-IoT / LoRaWAN backhaul · Built for existing infrastructure
The problem
Non-revenue water is the share of treated, pumped, paid-for water that never reaches a paying customer. Most utilities know the figure is high. Very few can say which street it is on. The instrumentation that would answer that question was never installed, and replacing every meter to get it is not a budget any council will approve.
Manual quarterly reading rounds produce one number per customer per season. By the time a strange consumption pattern appears on an invoice, the cause has been running for weeks and the dispute is unwinnable.
A background leak can run for a year underground. It becomes visible when it undermines a road or surfaces in a basement — at which point the repair is an emergency, at emergency cost.
Without segment-level flow and pressure, crews work from complaint maps and guesswork. Every dry hole is money spent, road reinstated, and public patience consumed for nothing.
Water drawn upstream of a meter never appears in any ledger. The only trace is a pressure and flow signature that nobody is currently recording.
The platform
Clamp-on and inline sensors fit the mechanical meters and junctions already in the ground. They stream pressure and flow continuously. Nabaflow learns each district's normal demand profile, then reports where live behaviour departs from it — with the segment, the evidence, and a confidence level attached.
Real demand falls near zero overnight. Loss does not. That difference is the signal.
Minimum night flow in DMA-03 has exceeded the expected band for six consecutive nights. Loss localised to segment SEG-14 between valve V-204 and district meter DM-07.
Sharp nightly pressure drop downstream of V-118 with no matching metered demand. Consistent with an unmetered draw-off; requires a field check before any conclusion.
DMA-01 inflow exceeds the sum of downstream customer meters by a widening margin over 14 days. Either a meter is under-registering or water is leaving the zone unbilled.
Node PS-2214 reporting 18 % remaining. Scheduled for the next maintenance round; telemetry unaffected.
Monthly meter read batch closed and exported to the billing system. Any node that failed validation is listed rather than estimated.
| Node | Type | Zone | Last seen | Battery | Status |
|---|---|---|---|---|---|
| PS-1041 | Pressure | DMA-01 | 4 min | OK | |
| FM-0326 | Flow | DMA-02 | 2 min | OK | |
| DM-0072 | Zone meter | DMA-03 | 6 min | Flagged | |
| PS-2214 | Pressure | DMA-04 | 9 min | Service | |
| FM-0411 | Flow | DMA-04 | 3 min | OK |
A node that stops reporting is an alert, not a silent gap in the record.
Nodes184
Night flow7.2 m³/h
BandWithin
Nodes96
Night flow18.4 m³/h
BandWatch
Nodes212
Night flow19.0 m³/h
BandAbove
Nodes140
Night flow5.8 m³/h
BandWithin
Figures shown are a worked example, not results from a live network.
Capabilities
Seven capabilities, all of them running on the same retrofit sensor layer and the same continuous data stream.
Clamp-on and inline units fit the mechanical meters and junctions already in the ground. No wholesale meter replacement programme, no waiting for a capital cycle.
Every node reports on an interval you set, between one and fifteen minutes. Pressure and flow together, because either one alone is ambiguous.
Anomaly models compare live behaviour against the learned profile of each zone and narrow a suspected loss to the pipe run between two known nodes — with the evidence attached.
Pressure drops that repeat on a schedule and do not match any metered demand are surfaced as candidates for a field check. Nabaflow reports the signature; the utility decides what it means.
Scheduled reads replace the manual round and export straight to the billing system. Failed validations are listed as failures rather than quietly estimated.
Campuses, towers and mixed-use blocks get consumption per unit, per riser and per tenant, plus burst alerts that reach a facilities phone at 03:00 rather than a caretaker at 08:00.
A standing water balance per zone: input volume, billed consumption, and the unexplained remainder — tracked against your own pre-deployment baseline, month over month.
How it works
A pilot normally runs on one or two district metered areas first, so the utility can judge the output before committing the rest of the network.
We map the district with your existing asset records, pick sensor positions at zone inlets, key valves and customer meters, and fit the hardware. Most fittings do not require a main shutdown.
Typically 1–2 weeksTwo to four weeks of continuous telemetry establish what normal looks like for each zone — weekday, weekend, seasonal and night-time. Alerting stays quiet until that profile is stable.
2–4 weeksLive flow and pressure are compared against the learned band. Departures are raised with a severity, a zone, a segment and a confidence level — never as a bare red dot.
ContinuousCrews work a ranked shortlist instead of a complaint map. What they find in the ground is logged back against the alert, which is how the model gets better on your network.
Ongoing loopPricing
Hardware is bought outright or leased. Monitoring is charged per active node. The analytics tier is what utilities add when they need a defensible water balance rather than an alert feed.
Hardware
$120 per node, outright
Or $6 per node per month on a lease, which folds the hardware into opex and includes replacement.
Most pilots start here
$4 per node, per month
Continuous telemetry, the alert feed, zone dashboards and sensor health for every active node.
Utilities
$500 per month, from
For water departments and authorities running a loss-reduction programme across multiple districts.
Indicative list pricing. Volume, lease terms and installation are quoted per network — tell us the node count and we will put a real number against it.
Evidence
Nabaflow is opening its first pilot programme. We are building partnerships with municipal and village council water departments, building operators, and university water and environmental engineering groups. Until those pilots produce results, these figures stay empty — and when they fill, they will be measured against each network's own pre-deployment baseline.
Every chart, alert, table and district figure on this page is a worked example built to show the shape of the output. None of it is drawn from a live customer network, and no utility, council or building operator named anywhere here is a current Nabaflow customer. Water infrastructure decisions are public-safety decisions; we would rather show you an empty field than a borrowed one.
Tell us the size of your network and how it is metered today. We will come back with a pilot scope, a node count and a real price — not a brochure.