How Long Before Anyone Actually Sees a Motion Alert at 2am?

The short answer: the alert takes seconds, the human takes hours

The technical part of a motion alert is fast — detection, capture, upload and push notification typically resolve while the event is still happening. The part nobody measures is what comes after: at 02:00 the phone is face-down, on Do Not Disturb, charging in another room, and the alert is read at breakfast. The realistic gap between a 2am PIR trigger and a human eye is not seconds or minutes — it is four to six hours, and on a weekend or a holiday it can be a day or more. That is why the important question is not "how fast is the notification?" but "what exists on the recorder when somebody finally looks?" A PIR camera answers with whatever clips its sensor agreed to record. An AOV camera — Always-On Video, an industry-generic term — answers with a continuous timeline covering the whole gap, with the alert acting as a bookmark into it rather than as the evidence itself.

This article breaks the delay into its actual stages, shows which of them you can measure yourself in about twenty minutes, and explains what a supplier can honestly promise and what they cannot.

The chain between an event and an eye

An alert is not one step. It is a sequence, and each link fails differently.

Stages 1 to 5 are engineering. Stage 6 is your phone's operating system doing exactly what you configured it to do. Stage 7 is biology and habit. Any specification argument that only optimises stages 1 to 5 is optimising the small end of the problem.

The stage nobody quotes: the person

Two behaviours dominate real deployments.

Sleep. Night-time alerts are, by design, suppressed by the phone. Modern sleep schedules silence notifications automatically, which is why people set them, and why the 02:00 alert waits until the alarm goes off. Nobody is going to change that policy for a camera.

Alert fatigue. Motion triggers do not only fire for intruders. Rain, moving branches, headlights sweeping a fence, a dog, an insect close to the lens — all of it looks like a heat-and-motion event. After enough of those, the rational response is to mute the app, widen the sensitivity threshold, or shrink the detection zone. Every one of those fixes makes the camera less likely to alert on the real thing. Alert fatigue is not a user error; it is the predictable outcome of a system that can only communicate by interrupting you.

The consequence is uncomfortable but simple: for unattended sites — a farm at 3am, a yard on a Sunday, a second property visited monthly — you should design as if no alert will be read until the next working morning, and judge the camera on what it recorded in the meantime.

What continuous recording puts into the gap

If the recorder is continuous, the four-to-six-hour delay stops being a loss. The timeline exists whether or not anyone was awake, so the morning review starts from footage rather than from an apology.

Note what changes about the alert itself. With continuous recording, the alert stops being the evidence and becomes an index entry: AI scene analysis flags a moment worth looking at, and you jump to that timestamp inside a recording that would have existed anyway. A missed notification then costs you response time, not the footage.

Measure your own latency in twenty minutes

Do not accept anyone's number for this, including ours. It depends on your cell coverage, your platform and your phone. Test it:

1. Baseline walk test, daylight. Note the wall-clock time you enter the frame. Compare it against three timestamps: the event time in the app, the time the notification arrived on the phone, and the first recorded frame that shows you. The difference between step one and the first recorded frame is your true detection loss.
2. Repeat with the phone in its night configuration. Sleep focus on, battery saver on, screen down. Do not silence the test to make the number look good — this is the configuration the camera will actually live with.
3. Repeat at the real mounting point. Signal three metres up on a pole beside a metal structure is not the signal you measured standing at the gate.
4. Repeat at night. Night imaging, colder scenes and lower supply all behave differently from a sunny afternoon test.
5. Approach slowly, and from the edge. Walk in at a crawl, and separately walk along the boundary of the detection zone rather than across it. This is where PIR triggering fails silently — and a failure here produces no notification at all, which is the result you most need to see before you buy.
6. Log it. Columns: test time, approach type, phone state, first recorded frame, notification received, difference. Ten rows tells you more than any datasheet.

Run this on a sample before a volume order. Samples ship in 7–10 days by air, which makes this an entirely practical step rather than a theoretical one.

What a supplier can honestly promise

Being precise about this is a good test of who you are dealing with.

Cannot be promised end-to-end: the delivery time of a push notification. It crosses a mobile network, a push service and a handset OS, none of which the camera manufacturer controls. Anyone quoting you a guaranteed seconds-level alert time for a 4G device is selling you a number they cannot enforce.

Can be promised, and should be in writing: that recording is continuous rather than event-only; the retention window in hours at your chosen resolution and frame rate on the specific SKU; that footage is written to local microSD, up to 128 GB as standard with larger cards on custom orders, so it survives a night when the network did not; and that video remains on the card when the upload fails rather than existing only in a cloud that was never reached.

The deeper mechanics of continuous recording at low standby power are covered in the AOV solar camera guide.

Where this changes the buying decision

For everything except the first row, retention beats latency. That is the whole argument in one table.

Specifying it

Our position on the hardware side, for buyers comparing quotes:

Browse the range on the products page, or look at dealer terms if you are reselling. For sizing on a specific site, retention figures for a specific SKU, or a sample to run the twenty-minute test on: WhatsApp +86 188 1869 9870 or info@solaguard.net. Quantities, prices and lead times always come from a person.