Commercial CCTV Installation
Commercial surveillance designed around risk and operations.
Plan and deploy selected enterprise CCTV environments with appropriate coverage, storage, access and resilience.
Discuss this requirement →Start with the environment, risk and required outcome.
Plan and deploy selected enterprise CCTV environments with appropriate coverage, storage, access and resilience. Namandla Consultants begins with evidence and recommends a proportionate engagement rather than a generic technology package.
Connected expertise for the wider environment.
Clear steps. Visible decisions.
What commercial CCTV involves that residential CCTV does not
The hardware overlaps. Almost nothing else does.
A four-camera system at a house is a self-contained appliance. A commercial system is a networked application with storage requirements, bandwidth requirements, retention obligations, access control, audit requirements and integration points into other systems. It is closer to a small data platform than to a security product, and the projects that go wrong usually go wrong because it was scoped as the latter.
Five things separate the two.
Pixel density has to be specified, not assumed. A residential system needs to show that something happened. A commercial system frequently needs footage that will support a disciplinary process, an insurance claim or a criminal matter. Those require identification-grade images at defined points, which is a calculation rather than a preference.
Storage is a sizing exercise
Retention periods are driven by policy and sometimes by regulation. Working backwards from a retention period to a storage requirement, across a realistic camera count at realistic bitrates, produces numbers that surprise people who have only priced the cameras.
The network is part of the system
Twenty-four cameras streaming continuously is a sustained load that does not behave like office traffic. Putting it on the same flat network as your ERP is a design decision with consequences, and it is usually made by accident.
Footage is personal information under POPIA
Recording identifiable people is processing. That brings notice, purpose limitation, retention, access control and operator obligations with it. Most installations in the market are silent on this.
Someone has to own it after handover
Cameras drift out of focus, hard drives fail quietly, firmware ages and the one person who knew the login leaves. A commercial system without a maintenance and monitoring arrangement degrades to roughly half its useful coverage within about two years.
Getting the design right
Pixel density, the number that decides whether the footage is usable
The international standard for video surveillance systems, IEC 62676-4, defines four operational levels by horizontal pixel density across the target. They are usually referred to as DORI.
| Level | Pixels per metre | What it supports |
|---|---|---|
| Detect | 25 px/m | Something is present and moving in the scene |
| Observe | 62.5 px/m | Characteristic details, clothing, direction of travel |
| Recognise | 125 px/m | A known individual can be identified with reasonable confidence |
| Identify | 250 px/m | An unknown individual can be identified from the footage alone |
The practical consequence is that a single camera cannot serve every purpose in a scene. A wide view of a loading bay at detect density tells you a vehicle arrived. It will not tell you who got out. Identification-grade coverage requires either a tighter field of view or a second camera positioned for the purpose.
This is where most commercial systems disappoint. The camera count was right, the resolution on the datasheet was high, and the field of view at each position was too wide for the density the purpose required. Establishing what each camera is actually for, before choosing lenses, is the whole design.
Storage sizing, worked properly
Storage is a function of bitrate, camera count, recording mode and retention. The arithmetic is straightforward and it is routinely skipped.
A 4 MP camera in H.265 at 15 frames per second in a moderately active scene runs at roughly 4 Mbps. Continuous recording, thirty-two cameras, thirty days:
4 Mbps ÷ 8 = 0.5 MB/s per camera
0.5 MB/s × 86,400 s = 43.2 GB per camera per day
43.2 GB × 32 cameras = 1,382 GB per day
1,382 GB × 30 days = 41.5 TB
Add RAID overhead and headroom and the requirement is closer to 55 TB usable. That is a materially different purchase from the appliance most quotations assume.
Three levers reduce it honestly.
Codec
H.265 gives roughly 40 to 50 percent reduction over H.264 at equivalent quality. Verify that every camera, the recorder and the client software all support it, because a single H.264 fallback in the chain removes the saving.
Recording mode
Motion-triggered or event-based recording cuts storage substantially in low-activity areas such as stores and plant rooms. It is the wrong choice on entrances, tills and cash-handling areas, where a gap in continuous coverage undermines the evidential value of the whole recording.
Smart codec and variable bitrate
Modern cameras reduce bitrate dynamically in static scenes. Useful, and worth setting a ceiling on, because a scene that becomes busy will consume what it is allowed to.
Network design
Surveillance traffic is sustained, unidirectional and intolerant of loss. Three decisions matter.
Segment it
A dedicated VLAN for cameras and recorders, with routing controlled at the firewall. Cameras are embedded devices with long patch cycles and they should not have a path to your business systems or to the internet unless a specific function requires it.
Size the uplinks
Thirty-two cameras at 4 Mbps is 128 Mbps sustained into the recorder. That is fine on a gigabit access switch and it is not fine sharing a 100 Mbps uplink with anything else. Aggregate at the switch feeding the recorder.
Power over Ethernet budget
A PoE switch has a total power budget as well as a per-port rating. Sixteen cameras at 12 W each is 192 W plus overhead, against switches commonly rated at 130 W or 180 W total. Cameras with heaters and IR illuminators draw considerably more at night, which is when they are needed and when the budget fails.
VMS or NVR
An NVR is an appliance. It records, it replays and it is bounded by its channel count and its disk bays. It suits single sites with a stable camera count and no integration requirement.
A video management system is software on a server. It scales across sites, supports failover, integrates with access control and alarm systems, and gives you role-based access, audit logging and export workflows. It costs more, it needs to be maintained and it is the right answer for multi-site estates and for anything with a governance requirement.
The decision point is usually not camera count. It is whether you need audited access to footage and integration with other systems. If you do, the appliance will not get you there and replacing it later means replacing the licences too.
What Namandla delivers
Site survey and purpose mapping
We walk the site and establish what each camera position is for, at what pixel density, under what lighting, at what times. That document drives every subsequent decision and it is what makes the quotation comparable to anyone else’s.
System design
Camera selection and lens calculation per position, recorder or VMS platform selection, storage sizing with the arithmetic shown, network and PoE design, and cable routing.
Installation
Structured cabling, camera mounting and alignment, recorder and server commissioning, network configuration and switch provisioning.
Configuration and tuning
Exposure, wide dynamic range, IR and white-light illumination, motion zones, privacy masking, retention schedules, user roles and export permissions. The tuning pass is done at the times of day the site actually operates, not at eleven in the morning when everything looks fine.
POPIA documentation
Notice signage wording, a processing record, a retention schedule, an access and export procedure and an operator agreement where a third party monitors the footage.
Integration
Access control, alarm panels, intercom, ANPR and building management where required. Camera call-up on an access event or an alarm is where a surveillance system stops being a passive recorder.
Ongoing management
Health monitoring across cameras, recorders and disks, firmware currency, quarterly image verification and storage capacity review, delivered under the same agreement as the rest of your IT estate.
Where we work
Corporate offices and campuses, warehouses and distribution centres, manufacturing and industrial sites, mining corporate and surface facilities, healthcare facilities, retail head offices and distribution, and multi-site estates requiring central viewing.
Onsite work is Gauteng. Design, configuration and system management run nationally, and for multi-province estates we design and manage centrally with contracted local installation at each site.
Where we do not work
Residential properties, complexes and small single-site installations of a handful of cameras. Those are well served by specialists who do them at volume and price them accordingly. We would be more expensive and no better.
POPIA and CCTV footage
This is the part of a commercial installation that most quotations do not mention, and it is a real obligation rather than a formality.
Footage of identifiable people is personal information
Where your organisation determines why and how that footage is processed, it is the responsible party for that processing under POPIA. Four practical controls follow.
Notice
Data subjects must be informed about the collection and purpose of the processing unless an applicable exception applies. In practice, clear entrance or area signage should be backed by a privacy notice that identifies the responsible party, purpose and contact route. The exact notice design should be confirmed by your information officer or legal adviser for the site.
Purpose limitation
Footage collected for site security cannot be repurposed for performance monitoring without a separate lawful basis and separate notice. This is where organisations most often drift out of compliance, usually with good intentions.
Retention
Personal information may not be retained longer than necessary for the purpose. An indefinite retention setting is a compliance problem as well as a storage cost. Set a defined period, document why and let the system enforce it.
Security and access
Access to footage must be protected by appropriate technical and organisational measures. For a commercial system that normally means named access, least privilege and an audit trail for viewing or export where the platform supports it. Shared administrator credentials and an unsecured recorder undermine that control.
Where a third party monitors your cameras, they are an operator under the Act and you need a written agreement covering their processing. Most off-site monitoring arrangements in the market run without one.
We produce the documentation set alongside the installation. It is not legal advice and it does not replace your own compliance review, but it means the system arrives with the operational controls already configured rather than as a project someone has to retrofit.
Common questions about Commercial CCTV
How much does commercial CCTV installation cost?
Commercial CCTV is priced per camera position plus infrastructure, and the infrastructure is usually the larger number. Cameras, cabling, PoE switching, recording platform, storage and configuration all sit in the total, with storage often the single biggest line on longer retention periods.
The variables that move a quotation are camera count and type, the pixel density required at each position, cable runs and containment, retention period, whether a VMS or an appliance is required and whether integration is in scope. Two quotations for the same camera count can differ by a factor of three entirely legitimately, because one sized storage for thirty days and the other for seven.
How many cameras do we need?
Camera count follows from purpose, not from site size. Establish what each position must achieve, at detect, observe, recognise or identify density, and the count follows. Sites are commonly over-camera’d in open areas and under-camera’d at the points that matter.
The positions that almost always need identification-grade coverage are pedestrian entrances, reception, goods receiving, dispatch, cash-handling points and server room doors. Wide-area coverage of yards, car parks and production floors is usually detect or observe density, which needs far fewer cameras than people expect.
What resolution do we need?
Resolution alone does not determine image quality. Pixel density across the target does, which is resolution divided by the width of the scene the camera covers. An 8 MP camera covering a twenty metre wide yard delivers lower usable density than a 2 MP camera covering a two metre doorway.
Higher resolution also costs storage and bandwidth proportionally, and it needs more light to produce a usable image because the individual sensor pixels are smaller. A well-chosen 4 MP camera with the right lens outperforms a badly positioned 8 MP camera in almost every real scene. Specify the density you need and let that select the camera.
How long should we keep footage?
Long enough to meet the stated purpose and no longer than necessary. POPIA does not prescribe one universal CCTV retention period. Many organisations use an operational baseline such as 30 days, while insurance, contractual or sector requirements may justify a different period. Document the basis and enforce it consistently.
Longer retention is a storage cost and a compliance exposure at the same time. Where a specific area genuinely needs longer, the better design is a differentiated retention policy, with most cameras at thirty days and a defined subset at ninety, rather than pushing the whole system out.
Can we view the cameras remotely?
Yes. Remote viewing is standard on any current platform, through a client application or a browser. The design question is how that access is secured rather than whether it is possible.
The correct arrangement is access through a VPN or a vendor cloud connector with multi-factor authentication, with the recorder itself never exposed directly to the internet. Port-forwarding a recorder to a public IP address is still common and it makes the system reachable by anyone who scans for it. If your current system was set up that way, it is worth changing regardless of who maintains it.
Does CCTV comply with POPIA?
A CCTV system is not compliant or non-compliant in itself. How you operate it determines that. You need notice signage, a defined purpose, a retention period, controlled and logged access to footage and a written agreement with anyone who monitors it on your behalf.
We supply the documentation set with the installation: signage wording, processing record, retention schedule and an access and export procedure. That covers the operational side. Whether your overall processing is lawful is a question for your information officer, and we will give them a clear description of what the system does so they can answer it.
Can CCTV integrate with our access control?
Yes, and it is one of the more useful things a commercial system does. An access event, whether granted or denied, calls up the associated camera and bookmarks the footage, so investigating an event takes seconds rather than scrolling.
Integration is straightforward where both systems support open protocols, ONVIF for video and a documented API or relay interface on the access side. It is difficult where either system is a closed proprietary platform, which is worth checking before you buy rather than after. We handle this on /biometric-access/ as well, since the two are usually specified together.
What maintenance does a CCTV system need?
Quarterly image verification, storage capacity review, firmware currency, disk health monitoring and physical cleaning of housings and domes. Without it, coverage degrades quietly and you find out during an incident.
The failures are unspectacular and cumulative. A dome that has hazed over, a camera knocked out of alignment by a forklift, a disk that dropped out of the array eight months ago, a firmware version with a known defect. Health monitoring catches most of it automatically. The image verification pass catches the rest, and it has to be a person looking at every camera because a system reporting healthy will happily record a wall.
Do you monitor cameras off-site?
We manage the system, meaning health, configuration, storage and access governance. Live off-site monitoring by an operator is a different service delivered by armed response and monitoring companies, and we integrate with the one you use rather than replacing them.
That distinction matters commercially. Managing the platform and watching the screens are separate disciplines with separate cost structures. Where a monitoring company is involved, we make sure the technical integration works and that the operator agreement required under POPIA is in place.
Can we use our existing cameras?
Often, yes. Any camera supporting ONVIF Profile S can usually be brought into a new recording platform, which protects the investment in cabling and mounting even where the cameras themselves are dated.
The assessment is worth doing before replacing anything. Cabling and containment are typically the largest cost in an installation and they usually stay. Where cameras are analogue, encoders can bring them onto an IP platform, though at some point the image quality makes replacement the better spend. We will tell you which of your existing positions are worth keeping and which are not.
How long does an installation take?
A thirty-camera single-site installation typically runs two to three weeks from order, with roughly a week of that on site. Multi-site estates and installations requiring new containment run longer.
The schedule is usually driven by cabling rather than by equipment. Where containment exists and routes are accessible, installation moves quickly. Where cable has to be run through occupied office space, a live warehouse or an industrial environment with access restrictions, the survey establishes a realistic schedule rather than an optimistic one.
When is a full commercial system the wrong answer?
When you have a single small site, a handful of positions and no governance or integration requirement. An off-the-shelf system from a specialist installer will cost a fraction of what a designed commercial installation costs and it will do the job.
The threshold is roughly where any of these become true: more than one site, a retention obligation you have to evidence, a requirement to integrate with access control, an audit requirement on who views footage, or footage that may need to support a formal process. Below that, buy the simpler thing. We will tell you if that is where you sit.