Monitored Site Security Explained: What Happens When a Camera Detects an Intruder

It's 2am on a Sunday. Someone cuts through the temporary fence behind your laydown area and walks toward the tool container. On a recording-only system, you find out on Monday. On a monitored system, a chain of events starts while they're still inside the fence.

This guide explains what happens when monitored cameras detect an intruder, one step at a time: AI detection, an operator's check, the siren and strobe, the alert to your phone, and the call to police. It also covers the types of monitoring on offer, where guards and patrols fit, what drives the cost, and how to set a site up so the chain holds together.

What happens when monitored cameras detect an intruder: the five steps

Every monitored site runs the same basic sequence. The hardware and the people change from provider to provider, but the order doesn't. Each step hands off to the next, and a gap anywhere gives the intruder more time.

Steps 3, 4 and 5 often overlap. An operator can be warning the intruder over the speaker while the alert is already on your phone and the police call is being made. The rest of this guide takes each step in turn, then looks at the choices you make before any of it happens.

Step 1: How the camera decides something is worth flagging

Detection is where most false alarms start, and where most of them can be stopped. Older systems used plain motion detection: any change in the picture triggers an event. On a construction site at night that means swaying shade cloth, headlights on the road, rain, possums and the odd kangaroo, all treated the same as a person with bolt cutters.

Modern site cameras use AI classification to tell people and vehicles apart from everything else. Siteguard's units use person and vehicle detection rather than plain motion, and detection zones are managed from Siteguard Foreman, so only the areas nobody should be in after hours are armed. For more on the analytics themselves, see how AI detection works on a CCTV tower.

Two set-ups do most of the work on a site:

• Line crossing: a virtual tripwire along a fence, gate or edge that triggers only when a person or vehicle crosses it, in the direction you choose. Our guide to line crossing detection covers where it beats motion detection.
• Intrusion zones: an area such as a laydown yard or plant compound that triggers when someone enters it. See how intrusion detection zones are set up on a construction site.

Schedules matter as much as zones. A zone armed at 5pm when trades work until 6:30 will flood the control room with your own crew. Match the armed hours to the site's real knock-off, and change them when the program changes.

The hardware sets the limits. On Siteguard's solar camera towers, the Twin Cam covers 180° with smart alerts, the Single Cam gives a 4K 8MP view across 95° for a single gate or asset, and the Pan & Zoom 360° Cam can be steered remotely with zoom up to 1km to check a figure at the far fence.

Step 2: An operator checks the alarm

Once an event fires, a monitored system puts it in front of a person. This is the step that separates monitoring from a camera that sends alerts to a phone nobody looks at after midnight.

With Siteguard's 24/7 control room monitoring, every alarm is verified by a real operator in an Australian-based control room before any action is taken. The operator looks at the live view and the clip that triggered the event and makes a call: a worker who came back for their phone, a kangaroo on the haul road, or someone who has no business being there. That cover runs on weekends and public holidays too, which is when most sites sit empty.

Verification does two jobs. It stops false alarm fatigue, the point where site managers mute alerts because most of them are nothing. And it means that when police get a call, it's about a confirmed person on site, with a live description, not an unexplained alarm.

The operator can only be as good as the information they have. Give the control room a current call list, the site's working hours and any expected after-hours visitors, such as a 4am concrete pour or a security contractor doing a lock-up check. An operator who knows a truck is due won't treat it as a break-in.

Step 3: Siren, strobe and a voice on the speaker

Once the operator confirms a threat, the aim is to make the intruder leave before anything leaves with them. Most people who climb a site fence are counting on not being seen. A siren and a flashing strobe on a tower above them changes that fast.

Siteguard's operators issue audio warnings and activate the siren to deter intruders before damage occurs. A live voice carries more weight than a recorded tone because it's specific: "You by the container in the grey hoodie, you're on camera and police are being called." When a voice beats a siren, and when it doesn't, is covered in two-way audio on security camera towers.

Light does part of the job too. White deterrence lighting and strobes make a person feel exposed and give the cameras a cleaner image to work with. For how the options compare, see sirens, strobes and lights on CCTV towers.

Deterrence works best early. A tower near the boundary that reacts as someone reaches the fence gives them a reason to turn around before the tool container, not after it's open. For the wider set of measures, see how to prevent theft on a construction site.

Step 4: The alert lands on your phone

While the operator deals with the intruder, the site contact hears about it. Siteguard's alerting system texts you a clear description of each event with a screenshot, so you know what's happening without opening an app or scrubbing through footage. What goes into one of those messages, and who should get it, is covered in how SMS alerts with screenshots work.

Decide in advance who receives alerts and what they're expected to do. A project manager who gets a text at 2am mostly needs to know the operator has it in hand. The person who has to meet police at the gate, or let a glazier in to board up a smashed window, needs to be on the list as well, with a backup name behind them.

The same events land in Siteguard Foreman, the browser-based dashboard that holds every camera, alert and recording across your sites. Its event linkage sets what happens when an event triggers: siren, strobe, relay outputs, push alerts, or all of them. Playback runs in the browser without downloading files, and the next morning you can ask Foreman "What happened on site last night?" instead of scrolling through hours of video.

Step 5: When the police get called

If the intruder ignores the warnings, or the operator can see a crime under way, the next step is escalation. Siteguard's control room escalates directly to the relevant authorities when it's required.

The timing of that call matters. Victoria Police, for example, asks people to call Triple Zero (000) to report a crime in progress or when the offender is still on site, and to use the Police Assistance Line (131 444) or online reporting once the offender has gone. A monitored site has an advantage here, because the operator can see whether the person is still there.

Whether police attend, and how fast, is their decision and depends on what else is happening that night. No camera company controls it, so be wary of anyone who promises a police response time. What the control room can do is give police a clear picture: how many people, where on site, what they're carrying and which way they left.

After the event, footage is evidence. Keep the clips, the alert log and any plate captures together. On sites with a gate camera, number plate recognition reads and time-stamps every vehicle in and out, which can matter more to an investigation than a shot of a hooded figure at 30 metres.

Self-monitoring, alarm monitoring and live video monitoring compared

"Monitored" covers several different services, and they don't respond the same way. Know which one is on the quote before you compare prices or coverage.

Back-to-base is the one most people know from home and shop alarms. It's a good fit for a building with doors, windows and power. The comparison with camera monitoring is covered in back-to-base alarm monitoring and how it differs from monitored site cameras.

Virtual guarding sits at the top end. Operators don't just wait for alarms; they look in on the site at set times, the way a guard walks a route. Our guide to virtual security guards for construction sites explains how it compares with a guard on site.

If your site is an open block with a fence, no building and no mains power, an alarm panel quote is the wrong quote. You need cameras that can be watched.

Where guards and mobile patrols still fit

Cameras and operators don't replace every job a guard does. A guard can check that a door is locked, meet a delivery at 5am, manage a crowd at an event and give evidence in person. Some contracts and insurers still ask for a guard on site during set periods.

The catch is cost at the hours sites need cover. Under the Security Services Industry Award pay guide, a full-time level 1 security officer's minimum rate from 1 July 2026 is $28.42 an hour for ordinary hours, rising to $56.84 on a Sunday and $71.05 on a public holiday. A casual officer starts at $35.53 an hour. Those are award minimums, before the security company adds its own costs and margin.

That's why many sites run monitored cameras as the baseline and bring people in only when the cameras show they're needed. A mobile patrol sent to a verified event does more than one that drives past on a fixed timetable. See security cameras vs security guards on construction sites for the side-by-side, and combining CCTV with mobile patrols for how a hybrid setup works. Guards are also one of the six layers in our guide on how to secure a construction site.

What monitored site security costs

There's no single price, and any figure given without seeing your site is a guess. The cost of monitored site security comes down to a handful of cost drivers:

• Cameras and towers: more zones mean more events to verify.
• Monitored hours: after hours only, weekends, or 24/7 including public holidays.
• Response level: alarm verification only, or scheduled checks and patrol dispatch as well.
• Hardware: hire or purchase, and the length of the hire.
• Connectivity: 4G where the signal is good, Starlink where it isn't.
• Site conditions: how many false triggers the site throws up (trees, roads, wildlife) and how much zone setup it needs.

Two questions come up most often. For guard pricing, see how much a construction site security guard costs. For the monitoring fee itself, see what site camera monitoring costs per month. The hardware side is covered in what it costs to hire a site security camera, and the full range is on the site security camera hire page.

Compare like with like. A low monthly fee for alert-only monitoring isn't the same product as an operator who verifies, warns and calls police. Ask each provider to describe in writing what happens at steps 2, 3 and 5, and weigh that against the cost of one unchecked break-in: stolen tools, an insurance excess, re-ordering and a crew standing around.

Setting up your site so the response chain works

The five steps only work if the site is set up for them. When monitoring lets a site down, the cause is far more often the setup than the operator. Work through this list before the first night:

1. Place towers where intruders come in: gates, the back fence, the laydown area and the site office.
2. Draw zones and lines around the areas nobody should be in after hours, and leave footpaths and roads out of them.
3. Set armed hours to the real knock-off, and update them for overtime, night works and the Christmas shutdown.
4. Give the control room a call list in order, with a backup for every name.
5. Flag expected visitors, such as an early pour, a crane delivery or a contractor's lock-up check.
6. Test the signal where the unit will stand. Where 3G and 4G don't reach, Starlink connectivity supports live streaming, remote camera access and real-time alerts.
7. Put up signage saying the site is under camera surveillance.

Power is part of the chain too. On a solar unit, the siren, strobe and lights draw from the same battery as the cameras and modem, so a tower that flashes at every passing kangaroo runs its reserve down. That balance is explained in how solar CCTV towers and trailers work. For sites with nothing on them at all, see securing a remote site with no power, internet or people.

Questions to ask a monitoring provider

Before you sign, get straight answers to these. A provider that can't walk you through the response chain step by step is selling cameras, not monitoring.

• Is every alarm checked by a person before anything happens, and where is that person based?
• What does the operator do at each step: siren, audio warning, call to the site contact, call to police?
• Who gets the alert, what does it say, and does it include a screenshot?
• Are nights, weekends and public holidays covered at the same level?
• Is the monitoring centre graded? ASIAL's grading scheme audits centres against AS 2201.2, covering construction, power supply, communications, operating procedures and redundancy, and publishes a list of graded centres.
• What does the event log or report look like after an incident, and how quickly can you get footage for police or an insurer?
• Who sets the detection zones and schedules, and who changes them when the site changes?

Write the answers down and keep them with the contract. When something goes wrong at 3am, you'll want to know what was promised at each handover point, not what you remember from the sales call.

Frequently asked questions