How Much Data Does an AOV Camera Use Per Month? A Practical Breakdown

The short answer

An AOV camera recording 24/7 to its own microSD card uses no mobile data at all for the recording itself. Continuous video is written locally; the 4G SIM is only there so you can watch live, receive alerts, and pull clips off the card from somewhere else. In practice one camera under normal supervision — a few short live checks a day plus AI notifications — lands somewhere around 1 to 3 GB per month. Check in constantly on the full-quality stream and you can push a single camera past 10 GB. Never open the app and the same camera might use a few hundred megabytes a month and still have every hour of the last two weeks on the card waiting for you.

That distinction is the whole subject. Almost everyone who asks "how much data does an always-on camera use" is quietly assuming that always-on means always-uploading. It does not, and the difference between those two architectures is the difference between a plan you barely notice and a bill that makes the camera unviable.

Recording and transmitting are two separate things

There are two things a solar 4G camera can do with video, and they have nothing to do with each other.

Recording is the camera writing video to the microSD card in its own body. AOV — Always-On Video, an industry-generic term, not one brand's technology — means the camera never sleeps between motion events; it runs continuously and uses on-device AI scene analysis to decide when to encode at full quality and when to hold a low-power baseline. All of that happens inside the housing. The SIM is not involved and the card fills at the same rate whether the SIM is active, out of credit, or physically absent.

Transmitting is you, elsewhere, asking the camera to send something. That is the only time the SIM does work, and it is entirely driven by your behaviour rather than by the camera's recording mode.

The practical consequence is worth stating plainly: a camera with a dead SIM still records. You lose remote viewing and alerts, not the footage. On a remote site that is a meaningful failure mode to understand, because it means a data outage does not create an evidence gap.

What actually consumes data

How to read these numbers

Treat those figures as planning estimates rather than specifications. Real usage moves with signal quality, how busy the scene is, how many people have the app open, and how long the app sits in the foreground after you stop looking. The only authoritative number is the one on your carrier's own counter after a full billing cycle.

A month of data, worked through

Three realistic profiles for one camera:

The two multipliers people miss

Two multipliers catch people out. First, the number of viewers: three managers each opening the live view is three streams, not one shared one. Second, notification volume in a busy scene — a camera pointed at a road or a working yard can generate a great many thumbnails a day, and while each is small, the total is not negligible.

Why AOV does not blow up the data bill

The intuition that continuous recording must mean continuous data is drawn from cloud-recorded CCTV, where every frame is uploaded and stored on a server. That architecture genuinely does consume enormous monthly volumes, and it is the reason so many people expect an always-on camera to be expensive to run.

Retention is a storage question, not a data question

Local recording is not a compromise made to save data. On an off-grid site it is the more robust design, because cellular coverage in a field, a yard, or a half-built compound is not guaranteed at every hour, and a recording architecture that depends on the link is an architecture with holes in it.

Storage is what bounds retention, not the data plan. microSD up to 128GB is standard, with larger capacities only on custom orders, so the question to settle at specification time is how many days back you need to be able to look — and to size the card and the recording quality against that, rather than against a number that sounds impressive.

Choosing and managing the SIM

• It must be a data SIM. Voice and SMS are irrelevant. What matters is data and, in many markets, whether the plan allows machine-to-machine use.
• Pooled multi-SIM plans win on multi-camera sites. Ten cameras each on a separate small allowance will have some idle and some over; one shared pool absorbs the variance.
• Check the band support against the carrier, not the country. Coverage maps describe where a network reaches, not which bands a device speaks. Confirm before you commit to volume.
• Watch the auto-renewal, not the balance. The commonest "the camera went offline" call is not a hardware fault; it is a prepaid data bundle that expired quietly.
• Signal quality costs data indirectly. A camera on a weak signal spends longer negotiating and retransmitting. If a position is marginal, fixing the antenna position is cheaper than raising the plan.

The SIM as a dealer margin line

For dealers, the SIM is also a margin question. Bundling a data plan with the hardware turns a one-time box sale into a recurring line, and it removes the single most common source of post-sale support calls — the customer who bought a voice SIM, or no SIM at all, and calls you when the app cannot connect.

Cutting usage without losing coverage

If a plan is genuinely tight, these are the levers, in the order that costs you the least:

1. Review on the sub-stream, escalate to full quality only when needed. This alone typically halves a heavy user's month.
2. Tighten AI detection to the classes you care about. A camera alerting on humans and vehicles sends far fewer thumbnails than one alerting on everything that moves.
3. Set detection zones. Excluding a public road or a swaying tree line cuts notification volume sharply and improves the alerts you do get.
4. Do bulk playback on site. Pulling a long sequence off the card over the network is the single most expensive thing you can do. If you can reach the camera, read the card.
5. Do not lower the recording quality to save data. It saves none — recording never touched the network — and it costs you evidence. Recording quality is a storage decision.

Note the shape of that list: everything on it is about what you ask the camera to send, and nothing on it degrades what the camera captures. That is the practical benefit of separating the recording path from the transmission path.

Specifying for a fleet

For a multi-site deployment, budget data per viewer rather than per camera. A contractor with 12 cameras and one person who checks them each morning uses far less than a contractor with 4 cameras and six people with the app installed. Build the estimate from the second number.

Resolution across the standard range of 1080p to 2K makes very little difference to monthly data under this architecture, because what you send is a live sub-stream and occasional clips, not the whole recording. Choose resolution for identification distance and storage retention; choose the data plan for viewing habits. They are separate decisions and treating them as one is how fleets end up over-provisioned on both.

See the catalogue at /en/products/, volume terms at /en/wholesale/, or the full technical background in the AOV solar camera guide.

For model recommendations and quantity pricing, message the team on WhatsApp at +86 188 1869 9870 or email info@solaguard.net.