Yes it’s true – 90% of CCTV systems did not achieve their objective.
Why would in-depth studies of large scale CCTV deployments report that 90% of these deployments had little to no significant impact on crime levels?
One simple factor – a lack of understanding with regards to the surveillance objectives.
Crime Increased after CCTV was Installed.
For instance – one estate installed numerous CCTV cameras all along the roadways. Vehicle crime incidents reduced. Burglary increased, along with the residents fear of crime and feelings of being unsafe. No cameras covered the back of the residences where burglars would enter.
Is mounting and connecting the camera all it takes?
Does Simply putting up 4 cameras around your home, make your home less susceptible to crime?
Not without understanding the objectives of surveillance.
Surveillance Objectives – DORI (Detect, Observe, recognise and Identify)?
Detect and Observe
Wide angled short range cameras installed. Who is watching the footage? Who is notified of a detection, enabling the monitoring of the intrusion? What response capability does the observer possess?
90% of home CCTV installations consist of cheap wide angled, short range CCTV cameras. This specification is intended for live viewing to detect and observe an intruder. Very few homes nor business premises, have a dedicated operator monitoring the footage 24/7.
In this instance we expect the camera to detect the intruder and then alert us.
Two questions arise. How accurate is the detection, and what response capacity does the person who is being notified possess?
Artificial Intelligence
90% of the cheap CCTV kits, and wide angled, short range cameras fitted to homes and business are not assisted by any form of intelligence. The detection alerts from these systems are so difficult to fine tune, that they are basically useless. Notifications will be sent out for everything from shadows, reflections of light, cloud cover, bright lights, bugs, dogs, cats, monkeys and many other nuisance events. Eventually they will be disabled due to the nuisance factor.
Of course Artificial Intelligence has not yet made such notifications fool proof. The nuisance factor is however greatly reduced by AI capable surveillance, and accuracy is increasing almost daily with the deployment of network protocol cameras. It is also possible to visually verify the cause of an alarm when a false alarm is received. In clean indoor environments, the detection is highly accurate.
Responding
Yet even when accurate notifications are received – what is the response capability? What interaction and level of co-ordination does the receiver have with emergency response teams?
Criminals can be in and out of a property in a staggeringly short space of time.
If your CCTV system is not enhanced with network protocol based intelligence, and not directly linked to a response resource , it becomes an incident investigation tool. Sufficient only to observe, after the fact – what has transpired.
Keep in mind that 90% of the deployed cameras are suitable only for detection – and now we will try to use them only for incident investigation. We sometimes even try to get evidential recordings from them.
Detection Cameras to Investigate
The cameras fail on both counts. Frustration mounts as you can watch the plunder of your home or business after the fact, but you cannot reliably determine what or who is responsible.
Another common scenario we see is wide angled, short range cameras deployed above the cash register. The surveillance objective is to investigate and reduce cash register shrinkage. Granted these cameras are rated to provide identification level surveillance at 4m.
In many instances the cameras are installed higher than 4m above the cash register. But even if the 4m range is maintained – the camera is rated to ID a person at 4M.
The problem is that our surveillance objective is not to ID a person at the cash register. Rather we want to detect sleight of hand and resolve fine notional characters, observe paper clips and other small counters placed beside the register. We hope to decipher the characters on the register screen, to detect colours and denominations of the coins and notes going in and out of the register. We also would like to integrate the video with our point of sales system.
Again, we are deploying detection and observation level surveillance for investigation and evidential recording objectives.
It isn’t surprising then to read a report stating that 90% of deployed CCTV installations failed to achieve their objective.
Do you have the time to review the footage?
CCTV cameras are installed to record activity that the occupant is not present to observe personally.
These recordings can be used for incident investigation, which provide data for prevention, and even corrective training.
The store manager isn’t able to observe the entire store, so CCTV is deployed. Have you considered how many hours of footage is generated by 16 cameras recording for 8 hours of the day?
Which manager has the time to sit down and review that amount of footage?
Once again the CCTV system loses it’s power as a managerial and proactive resource, and becomes a reactive resource used only in an emergent situation.
Smart Search and Smart Data
A CCTV system designed with industry specific intelligence – such as people counting and heat maps, or collision avoidance technology – makes intelligently collected data, quickly available to management. The data is used to pro-actively improve service and profitability, reduce losses and deter crime.
Modern systems are also equipped with smart search capability that allow quick data retrieval using a few criteria. Operators can quickly extract the required footage and save much time.
Active Surveillance becomes Passive
Another common scenario is the deployment of active surveillance cameras (also known as PTZ or Pan, Tilt and Zoom) without any permanent operator manning them. The chances of a PTZ camera detecting a crime taking place in front of it’s lens is remote at best. These cameras are incident investigation tools,meant to be complimented by fixed passive detection and monitoring cameras. Yet they are deployed to detect incidents?
Clearly it is expected that 90% of deployed CCTV systems are not going to achieve their desired objective. Especially when installed by the corner IT geek, the plumber, electrician or motor mechanic down the road.
Who is installing these defunct systems?
Are the technicians entirely to blame?
Why does a technician feel compelled to install a detection camera for an incident investigation task?
Often it’s the customer that applies price pressure.
The plumber or my corner computer sales person can do this for me a lot cheaper, he also does cameras you know?
A 2.8mm, 1 megapixel Camera with 20 meter night vision is after all the cheapest camera you can supply. Couple this with a 1080N (1 megapixel Recorder) and pretend it’s 1080P (2 megapixel) to supply an even cheaper system.
Operational Requirements
Alternatively, customers don’t feel compelled to invest the time required to walk through an operational requirements process. A process that clearly defines the surveillance objectives for every sphere of operation at the facility.
Crime Risk Analysis
Customers don’t feel compelled to invest the time to compile a crime risk analysis. A process that identifies the probability and consequence of specific criminal events. A process that allows targeted risks to be addressed, with budgets applied – relative to the expected ROI.
The customer usually prefers to take you on a rushed walk about the facility, pointing out where they would like a camera.
Yes it’s true – Your CCTV system has probably been a waste of money.
If you are going to buy a generic kit of cheap cameras and recorder and stick them up about your house – the unfortunate truth is that you are throwing away 90% of whatever you spend on that. It most likely won’t prevent a burglary, or home invasion – it won’t summon rapid assistance, and will not help you prosecute at all. Neither will it assist with intelligence, managerial assistance and business data.
