By the end of 2000, central Durban’s CCTV system had 67 cameras. Contemporary reporting said they had originally served traffic monitoring and were increasingly being used for crime prevention.
Sixty-seven is the kind of number that makes a strong headline. The operating detail behind it is more instructive.
Anthony Minnaar’s later academic review says funding and staffing constraints meant there were not enough operators for continuous live monitoring across the system. Some camera inputs were recorded and archived without an operator being able to react immediately to every event.
The distinction is simple but often overlooked: camera count measures installed endpoints; it does not measure operational capability.
Durban’s Safer City upgrade
In November 2000, a contemporary IOL report described a R2.5 million allocation for upgrading and improving the Eye-in-the-sky system, monitoring and training. It also reported a further R7.5 million over the following two years.
The wider Safer City approach combined effective policing, targeted social crime prevention and environmental design. The city planned to use research to identify serious-problem locations before deciding where technology was necessary. Proposed expansion included the CBD, Warwick Avenue, taxi ranks and townships.
That framing is more sophisticated than “put up more cameras.” It recognises that place, people, policing and design all influence what CCTV can achieve.
Four different CCTV counts
When evaluating a Durban or KZN system, at least four counts may be relevant:
- Installed cameras — physically present on the site.
- Connected cameras — reaching the recorder or video platform.
- Recording cameras — producing usable stored footage.
- Operationally monitored cameras — being viewed or surfaced to a person who can respond.
A fifth measure is even more important: how many cameras can deliver the required evidence when an incident is investigated?
One camera may be powered but out of focus. Another may record with the wrong time. A third may have a clear general view but no usable face detail. All three could appear “online” while failing their intended purpose.
Training was already part of the budget
The 2000 Durban plan explicitly included monitoring training. That is historically important because it shows the human layer was recognised early.
Operators need a clear escalation process, camera names that correspond to real locations, predictable playback controls and permission to export evidence. A system that requires one expert to remember every channel number and password is fragile.
For homes and small businesses, the “operator” may be the owner using a phone. The same principle applies. If notifications are overwhelming or playback is confusing, the system will not be used well.
The practical Durban lesson
Do not judge a CCTV proposal only by the line that says “16 cameras” or “32 cameras.” Ask what those cameras are expected to do and how the rest of the system supports them.
Durban’s 2000 milestone is valuable because it records both ambition and constraint. The city had moved beyond a small beachfront pilot, but the network’s growth created an operating challenge. Modern systems can automate more than their analogue predecessors, yet they still need intentional monitoring, recording, response and maintenance.
Research note
This article uses Minnaar’s 2007 academic case study and Paddy Harper’s 10 November 2000 report on Durban’s Safer City plan. Counts and budgets are historical, attributed figures and must not be read as current municipal data.