As a design rating, yes — that is what LPWAN is for. A AAA cell holds roughly a watt-hour, which across fourteen years leaves an average draw in the microamp range. It only works if the device sleeps almost continuously and wakes briefly to send a very small payload. It is an energy budget, not a measured field result.
Case study
Fourteen years on one AAA cell — and why that is a design decision, not a result
2017 · Russia and CIS · Public. Utility metering where the hardest requirement was that nobody should ever have to visit the device.
What it was
Telecan collected electricity, heat and water meter readings autonomously and pushed them to cloud analytics. No walking the building with a clipboard, no scheduled visits to read a number off a dial.
The architecture spanned sensors, the radio network, collection, storage and analysis. But every meaningful decision in it traces back to one constraint: a device installed in a basement, a riser or a meter cupboard has to survive there without maintenance for as long as the billing cycle cares about it. Field visits are the entire operating cost of this category. Remove them and the product works; keep them and it does not.
What the 14-year number actually is
The sensor was rated for 14 years on a single AAA cell.
That is a design specification, not a measurement, and the distinction matters enough to state plainly. The project ran in 2017 — a fourteen-year field result could not exist, then or now. Anyone claiming a measured fourteen-year life in 2017 would be claiming something arithmetically impossible.
So the honest version is more useful than the impressive one: this is an energy budget the design was engineered to meet, and the interesting question is not whether it was proven but what has to be true for it to hold.
How a rating like that is earned
An AAA cell holds roughly a watt-hour. Spread across fourteen years, the average draw available is on the order of microamps. Nothing in that budget can be spent casually.
Which means the radio dominates everything. The device is asleep almost all of the time, waking briefly to send a very small payload, and the ratio between those two states — the duty cycle — is what the whole rating rests on. A meter reading is a few bytes that change slowly, so the design can afford to be extremely sparing about when it speaks. That is the trade LPWAN exists to make: give up bandwidth and latency, buy years.
The exact transmission interval, device class and modulation are not published, so they are not stated here. The mechanism is sound; the parameters are not ours to invent.
The network around it
| Protocol | proprietary Lumiot LPWAN, with LoRa listed as planned |
| Devices per base station | up to 2,000 |
| Radius | up to 4 km |
| Gateway uplink | GSM |
| Resources metered | electricity, heat, water |
Those numbers describe a deliberate shape. Two thousand devices inside a four-kilometre radius is dense urban coverage from very few installation points — which matters, because base stations are the part of the system somebody does have to visit. The economics of the whole deployment come from that ratio.
The GSM uplink is the pragmatic half of the design: proprietary radio where it buys battery life, commodity cellular where coverage already exists.
What this case shows
| Claim | Evidence |
|---|---|
| The protocol was proprietary, not off-the-shelf | Lumiot LPWAN, with LoRa support listed as planned |
| The network shape is specified | up to 2,000 devices per base station within a 4 km radius, GSM uplink |
| Three utility types were covered | electricity, heat and water metering |
| The battery figure is a rating | 14 years on a single AAA cell as a design specification, not a field measurement |
What it does not claim
- Not a measured lifetime. Fourteen years is what the design was engineered to reach. No field result of that length exists.
- No transmission parameters. Interval, device class and modulation are not published, so the mechanism is described without inventing the numbers behind it.
- No sensor or interface list. The specific meter models and their physical or digital interfaces are not documented publicly.
- No deployment count. The platform was tested in Russian and CIS cities, but a verified number of installed meters is not available.
FAQ
Frequently asked questions
Calculated. The project ran in 2017, so a fourteen-year field measurement could not exist. It is the lifetime the design was engineered to meet.
A proprietary LPWAN protocol called Lumiot, with LoRa support listed as planned. Base stations supported up to 2,000 devices within roughly a 4 km radius, with a GSM uplink from the gateway.
In 2017 the trade was different than it is now. This design used proprietary radio where it bought battery life and coverage density, and commodity GSM for the gateway uplink where cellular coverage already existed — proprietary only where it earned something.
Up to 2,000, within a radius of about 4 km. That density is the point: base stations are the part of the system someone still has to install and service, so the fewer of them per covered meter, the better the economics.
Prefer a conversation first?
If you would rather talk it through before sending a brief, book a short routing call and we will point you to the right next step.