printer friendly version

City surveillance: 20 years later

Twenty years down the line and having completed many city surveillance systems and expansions thereof, we find ourselves questioning the technical, operational, legal and maintenance methodologies followed in southern Africa.

During the late ‘90s, we published a summary of a thesis on the utilisation of CCTV in public spaces (CCTV in the streets the true facts, N. Strauch, 1997), which was based on research we had done. The aim at the time was to identify what was done in other countries (Europe) and specifically to find out what would suit the African continent and to learn from their mistakes.

Well, 20 years later and the conclusion seems to be the same, more ‘new’ companies have joined this lucrative sector of the market, re-inventing this wheel many times with new technologies leading the operational output, untrained staff and low maintenance uptimes. The operational methodologies have gone a full circle where we find ourselves doing things the way they were done in the late 1990s.

Yes, it could be said that there is a success story behind each of the systems operated in our country and it is not my intention to criticise any of these. The experience gained over the years where city surveillance has remained my core interest needs to be told, if only to shed some light on what works and where the pitfalls are.

As a third-party company (a specialist subcontractor), we give our clients the full service ranging from the design of these systems, the civil works, fibre installation, planting of poles, control room layout, the installation of the cameras and head-end culminating in the handover of a functional system.

In saying this, we have had many frustrating discussions with end-users, consultants, clients and suppliers who deem their new systems as the solution to all a specific city’s problems. This does go a long way, but is generally seen primarily as a crime prevention tool. It has been difficult to alter this perception to the understanding that it is a city management tool.

Then again, managing the city by these means, it seems, could allow a service provider to step into the shoes of those responsible for other services such as electrical, water, sanitation and traffic; the reporting thereof could thus be harmful. On the other hand, in cities where the tool is embraced as a management tool, these managers have the facts at their fingertips. Water leaks are fixed immediately, rubbish heaps are cleared, traffic lights and traffic congestion is taken care of and streetlights are functional.

Electronic patrols of each area covered by the surveillance technology can be done frequently, picking up all management incidents and reporting these to the different departments.

Why is this then not the case?

Six main methodologies need to be understood, specified and implemented.

1. City surveillance is not about the cameras, the people or the technology used. City surveillance has only one output, the footage.

2. City surveillance is not only a crime-fighting tool, it is a city management tool.

3. Hot spotting where cameras are deployed at statistically high crime points throughout the city is a recipe that just does not give the successes hoped for. Each city has its own character, thus a design that enables track and trace by the operator brings success.

4. Operators are a unique breed of person. They are not uncles, cousins, children or a D grade security guard. Operators should have the ability to proactively identify the incident, follow it through its lifecycle and build the case.

5. Smart cities where technology warns of gatherings or loitering, rings bells when illegal parking or driving is analytically determined and criminals are identified anywhere in the city via facial recognition, might work in the minds of clinically calculated software design professors. South Africa, however, is undisciplined and electronic rules that are set-up to detect exceptions just cannot be implemented and are impractical to police.

6. Maintenance policy is critical.

The footage

It has become critical that attention be given to all factors that influence the quality of the footage during the design phase. With the flood of technology advancements and dumping of cheap products into the South African market, we seem to believe the specification sheets attached to tender documents. The IP products in particular have created a storm, some of the end users deem the words digital and IP to be this massive advancement and even believe that all IP products, by virtue of the name, are the same.

Bad footage, even when compared to old analogue recordings, are popping up in the market and when utilised in the public surveillance scenario will cause cases to be lost in a court of law.

Simply put, if the operator is correctly selected and the system is designed specifically for city management, then recording space is only needed for 48 hours and in some cases 72 hours. If the operator did not see the incident and the recordings are not reviewed, then only the odd, by chance incident recording could mean something. Any incident that was not seen by the operator should be picked up on the next rotation of the camera or from a static, and there will thus still be time to respond.

City management

Over the years, incidents have been logged, recorded and processed, giving a perception that crime is being tackled by the utilisation of city surveillance. Is this the case? Yes, but what about all the other service related incidents a city can be proud of?

Some cities that utilise the tool to manage all their service-related incidents should be screaming this success from the top of their highest buildings. During a normal day service delivery managers are faced with many challenges, some of which seem trivial, but to the citizen on the street it is noticeable when things are different.

Traffic lights seem to go off just before morning traffic and then there is chaos. How true is this, and all we do is sit there in queues hoping that someone will send out a crew to either repair the fault of have a points man deployed. In the world of an organised city where embracement and understanding of the added value of a surveillance system is prevalent, these failures are reported as and when they happen, with immediate reaction and repair even before the city awakens.

An interesting case that comes to mind that can demonstrate the power of effectiveness in both city management and the fight against crime happened when an operator who took her job seriously noticed a city owned truck had been parking at the same place every Wednesday. Initially, just mentally recording the fact, she made it her personal project to find out more.

During her break she then approached the review specialist and they reviewed and stored this footage. As time went by they both realised that there was something fishy going on and involved their managers, who escalated this to the responsible person in the city. A project was registered and all eyes were on the truck, every move was watched and eventually it was confirmed that diesel was been sold off to small users. The decision to hold back on arrests paid off and in the end, with buy in from the manager involved, the Metro Police and SAPS, a whole ring of criminals was apprehended and convicted.

Planned operations with Metro Police to enforce bylaws such as closing times and selling of alcohol at night clubs all under the watchful eyes of the control centre can be done. During such operations, having SAPS and Metro both in one vehicle attending to crime and bylaw enforcement at the same time, can only be dreamt of.

Some centres have allocated seats for city management functions, Metro Police and SAPS, but this is few and far between. There is unfortunately, in some cases, a sense of degrading the position of such officers especially where a service provider manages the centre. The choice to utilise own staff or to outsource is a discussion that will be had at another time.

These officers, if motivated and chosen correctly, can drive such a centre to its maximum. However, it would need a senior official to whom they would report who is dedicated to the centre and physically sits here. Even leading the Metro and SAPS vehicles to a crime scene via radio is problematic when done by a service-providers supervisor, because these services have protocols and don’t really like listening to a civilian.

Once again, the solutions are easy but need dialog and buy-in from the authorities in that city. In places where it has been implemented, it works reasonably well as long as the dedication and the chip on the shoulder of some is not the important factor. There are many more such management functions that could be discussed and it is merely the creative thinking and willingness to utilise the surveillance centres to their fullest that will make the difference.

Hot spotting

Evaluating a city during the design phase takes input from all involved to ensure that the most effective placement of cameras is determined. Years ago, we would look at a city to find what we then termed crime drawers. These are spots where criminals are naturally drawn to: ATMs, taxi ranks, banks, shopping centres, stations, taverns and main walkways are some examples. We would then evaluate statistics and overlay them on these spots to determine the factor of accuracy. This is called hot spotting.

As time moved and experience was gained, we found this to be ineffective. This mainly because the reaction teams had to race from spot to spot and perpetrators would inevitably be long gone. Printed snapshots of the perpetrators pinned to a wall showed no real success and they were never seen again. The advent of facial recognition software caused much excitement and product sales people crossed our doorstep on a daily basis. It was said that we could load this wonderful software and when a face which was loaded on the database was recognised, it would warn us and we could send the SAPS to apprehend him.

Little did we know that hit rate, lighting and the angle of view would have such an influence that to date this has not succeeded. Operators are there to build the case through its lifecycle, not to hit the reject button every 10 seconds because of mismatches.

The implementation of a successful city surveillance system is based on in-depth design, taking into account the crime drawers, but rather overlaying the city with three layers.

Layer one is the city entry and exit level. Here we identify the entry, to moving through, and the exit routes of the city. Typically cameras (PTZ domes and static) are placed along these routes. Number plate recognition can be utilised as an aid at this layer.

Layer two is the tracking layer. These are placed along secondary routes and are aimed at tracking vehicles and people along these routes.

Layer three is the trace layer. Generally, these cameras are placed at intersections, using them in all four directions, with a camera at the next intersection one street up. This causes a matrix effect covering the area with a video blanket.

It can be argued that this is almost impossible to achieve because of the quantity of positions needed, but it should also be noted that housing areas are not covered in this fashion, the entry / exit and the track level is utilised in these areas. It would be impossible to cover each house and even when a house is in close proximity, the first question is usually, “Can the camera see into my bedroom?” Yes. Please close your curtains if you don’t want to be seen.

Typically, a medium city such as Bloemfontein would need about 46 entry / exit domes with 184 statics, 93 trace dome cameras and 183 trace cameras. Costing of such a system, when done correctly with a UPS at each point, surge and lightning protection, quality power supplies and legal electrical terminations, would average at R380 000 per point, including the total fibre infrastructure, control mechanisms and control room.

A smaller city such as Klerksdorp would need about 38 entry / exit domes with 152 statics, 48 trace dome cameras and 42 trace cameras.

Operators

It is quite evident that operators are generally not selected or tested against any measured criteria. They are mostly employed by choice of management, or requested by clients – uncles, sons, children, and many tenders call for a D-grade guard. If this is the criteria, then no wonder the success of these multimillion Rand systems are not as successful as we want them to be. Models developed by professionals such as Dr Craig Donald, Steve Clupp and even some experienced centre managers just seem to fall by the wayside.

Many questions remain unanswered and this mainly because the funding is not available and the fact that service providers cannot afford the services of a professional. By adding a fixed amount, a line item in tender documents that can be used to pay for the services of such a professional would enable and force the selection of this special breed of person.

Technology is moving at a rate that has become far superior to the abilities of the R5 000 a month person, and in cases where skilled people are employed the task at hand gets boring (because of the lack of utilising the system to its maximum extent) and they only last till the next job comes up.

Smart cities and technology

We have come a long way since the days of VCRs, PTZ cameras with 300 mm lenses, housed in gigantic plastic bubbles with no optical correctness, IR lighting the size of a A4 sheet of paper that no one could afford and fibre cable with mechanical splicing. Our belief in the technology has remained and developments like these have changed the way we do things.

The only constant is that we live in an undisciplined South Africa where rules are made to be broken and are impossible for the law agencies to police. Jaywalking is the norm, not the exception; crowds gathering and walking on the inner roadside around parked cars is a minute-to-minute, second-to-second event. Double parking and taxi drop-offs are anywhere and the litter boxes along the streets are there to serve as a seat or a street vendor’s table. How then can we take a disciplined, analytical approach in an attempt to aid the operator to identify these exceptions when they are the norm?

The utilisation of analytics is not totally impossible in the city scenario and has been used successfully to warn of pattern changes it traffic, double parking in busy main arteries, one way traffic violations and even where vehicles have crossed the centre on double ways. The challenge arose in policing these incidents, the Broken Window Syndrome starting with management who dictated policy on what should be reacted to, gradually watering the use thereof down to disuse and eventual switch-off.

Good planning, great management, professional operators and massive buy in is needed to overcome these challenges. It can be and is done to a certain extent at some of the centres, but the over selling of functions within the technological space is being interpreted as the full truth, leading decision makers to believe that they can get away with less operators when using this wonder drug called technology.

Latency, for example, is an inherent factor within the IP world we live in. There are some schools of thought that argue that it has no effect on the way that operations are carried out because the operator gets a feel for it and can compensate whilst building the critical case. Is this true? It might be dependent on what that latency is?

Maintenance policy

Maintenance is the aspect that is generally misunderstood or underestimated. A maintenance policy should define the acceptable levels of uptime ensuring that the system is always ready for use. Maintenance is defined as preventative and corrective maintenance, but what does this mean and how can it be measured?

It is important to realise the depth at which maintenance should be measured. Strategically maintenance, support and technical sustainability is a disciplined, unified and iterative approach to those management and technical activities needed to:

• Acquire the required support. During the design phase we should ensure that the maintenance elements such as personnel, tools, spares, vehicles and workspace is included.

• Provide the required support during the operations at minimum cost. Having the correct personnel and tools available for the task at hand.

Maintenance tasks are grouped as follows:

Corrective maintenance tasks restore a failed item. This is easily done if the design caters for fast and effective replacement of these items. An example thereof is our patented STEVE (Surge Technology for Electrical and Video Equipment). This product is mounted within the pole structure and carries all power (including UPS) and surge protection elements for the cameras in the field. Diagnostics is easy and the replacement of the STEVE is via plug-in connectors. Low-level technicians can thus unplug the unit and replace it with a spare within three minutes ensuring that the downtime is minimal.

Preventative maintenance tasks systematically inspect, detect and correct incipient failures, primarily wear-out failures, either before they occur or before they develop into major failures. Simple test sheets are utilised to measure all Voltages, Amps and Ohms and log these into the database. It is surprising how soon the naughty boys (camera locations and control room elements) are identified via graphs aiding in the replacement of items which are not performing to specification. The added value is empowering the technical team to use facts about failed items during the warrantee period, identifying possible bad batches or component / design failures within a product. It has been our experience that the product suppliers are grateful and work together with us to resolve inherent product glitches.

The following are examples of the determined policy statements.

• Operating Time (Mission) = Daily operations of the system is to be 24 hours per day.

• Peak Operational Times = Between 05h00 to 08h30 and 15h00 to 18h00 daily including weekends and public holidays.

• Minor adjustments may be performed during operations, but no adjustment may be performed during peak operational times.

• No preventative maintenance shall be performed during peak operational times, only urgent operational repair, limited to the replacement of LRUs will be permitted.

• Failed LRUs will either be discarded or sent to intermediate support for repair.

• Only general purpose or standard test and support equipment will be used at organisational level support.

• Tests carried out at organisational level will be logged for intermediate and depot use.

• Replacement units shall be plug-in and require a minimum amount of fasteners, or in the case where special fastening methods are to be used for vandalism purposes, all fasteners are to be standardised.

By utilising such a policy, dictated upfront during the tender and included in the design, a stable system can be guaranteed. It is no use blaming Eskom when we don’t have enough power backup. We can’t pass the buck when the 10 cent power supplies that we used keep failing.

City surveillance is a specialist field which has taken us 27 systems throughout Africa to understand. We have the luxury of not being a product supplier, thus giving our clients the best advice available without the pressure of having to adhere to manufacturers' sales targets and agency agreements.

A city surveillance system has to function 24/7, why then do we deem an aircraft more important when it only performs its purpose 20% of the time.

Share this article:

Further reading: