A cost effective surveillance system needs to operate efficiently ensuring that all components are operational and adaptive to its environment and that minimum intervention is required by the user.
Self Diagnosis & Health Check
In any large network of cameras, on average 20 % will not be operational at any time. However, the user usually does not know which ones are not working until there is an event.
The iOmniscient Health Check system displays the entire system on a dynamic map. The system indicates if a camera is not operational for a number of reasons which can include the simpler applications of detecting tampering or sabotage but also more common but complex problems such as the camera having lost focus or having been moved due to vibrations. The system even provides a warning if the cameras are working but the images cannot be seen due to bad weather or if the cameras require cleaning.
NAMS - Nuisance Alarm Minimization System
The minimization of false alarms is critical for any video based system. Every iOmniscient product is armed with an Artificial Intelligence based NAMS module. In a recent system test against one of iOmniscient's major US competitors which lasted over a week, both sides achieved similar detection accuracies. However, iOmniscient was differentiated by its ability to cope with the false alarms. The competitor was getting 200 false alarms each night compared to zero from iOmniscient.
Of course, it is not possible to eliminate every single false alarm and humans themselves can often be fooled by light changes and mirages. However, a good NAMS module is critical for any decent Video Analytics system.
Smart City applications can involve thousands of cameras and other sensors. The hardware design must be able to allow the system to scale from a few cameras to thousands. A federated structure is useful. The system on a railway line provides a good example of a federated structure. At each station, the station master must receive sufficient information to run his operation effectively. Management at the central control room must also receive more summarized information than they would require to manage the operation at their level. The system must be able to provide the right level of information for each level of management.
Maintaining a very large, distributed network of computers and computing devices can be difficult and expensive if it has not been designed well. Many edge devices cannot be managed or maintained remotely. All iOmniscient systems are designed to be managed, diagnosed and maintained remotely independent of the architecture.
A comprehensive Smart City system must be designed with a totally open architecture which means that the user can choose any supplier for the cameras and for other associated hardware. The user must be free to purchase cameras and computers from any supplier as long as the hardware meets certain minimum specifications and as it is appropriate for the task.
The investment in a Smart City system must be a long term one. The component that is needed to be replaced first is the hardware as the software can evolve and be upgraded. However, the hardware will become obsolete or will wear out within a few years. The software must therefore be designed to work not only with the hardware that exists today but also with the hardware that may be selected in five or ten years. Therefore, it is imperative that the software is NOT proprietary. It must be totally open to ensure that the best hardware can always be used.
The majority of Smart City applications will occur in complex scenes with constantly changing environments. iOmniscient's systems have been designed to continuously learn, understand and adjust to the changes in their environment, for example, to light changes and variations caused by slow moving shadows. They can thus cope seamlessly with transitions from cloudy to sunny periods, from day to night and back to day. The system is able to perform this adjustment without manual configuration.
Maximizing your Return on Investment
A Smart City System is a major investment for stakeholders. There are several factors that affect the maximization of the return.
The normal assumption is that minimizing the cost of the system is the best way of maximizing the return. Unfortunately, getting the cheapest system usually results in the implementation of a system that does not meet the objectives of stakeholders. It is better to have no system than to have spent a significant amount of money on a system that is ineffective.
Minimizing Cost by Design
And an effective system does not have to be more expensive. It is only more expensive if it is not well designed.
For instance, as we saw in the chapter on Automated Surveillance, the use of iQ-Hawk can reduce the cost of storage and network bandwidth by such a significant amount that even after adding the cost of the software, the overall cost of the solution would be much lower. Of course, to achieve this benefit, the whole system has to be designed taking into account the advanced capabilities of iQ-Hawk. If the system is designed using conventional assumptions, these cost reduction benefits are not realized.
Therefore, the lowest cost software may not result in the lowest cost system. The overall cost of the system is important rather than the cost of any component.
Other architectural assumptions can also affect the cost of the system. Systems are often designed to be Centralized or Decentralized which is usually based on the technology provided by the vendor. If the vendor provides an Edge based solution, there is a propensity to design a decentralized solution and vice versa. Since the iOmniscient solution is available both in a server based centralized architecture and in a Super Edge based decentralized one, the actual design can be most appropriate to meet the objectives of the stakeholders. The most cost effective systems are hybrid systems, such as those available from iOmniscient, as they provide the flexibility for different user groups to use different architecture to meet their objectives in the most cost effective manner without compromising the overall design of the system.
As we have seen, some tasks are better performed centrally and others work better when distributed. If the architecture is not the constraining factor, the overall cost can also be minimized.
Irrespective of the architecture, the level of intelligence available on the system has a drastic impact on the effectiveness of the system. A system that is incapable of working in a crowded system will merely give an unacceptably high number of false alarms in that type of situation. Selecting software with the right iQ level for the task to be performed is the most critical factor in determining the value of the system.
Having the right Cameras and other Infrastructure
Users often spend significant amounts of money on cameras which may not be appropriate for the job at hand and which may be placed in positions that are not appropriate either. There are numerous examples of cameras being installed with inappropriate views. Cameras for counting systems have been installed outside of elevator doors where people wait causing false counts. Moving the camera a short distance away provides a perfect counting view. Cameras have been installed for License Plate Recognition even when they do not have an adequate shutter speed and when they have inappropriate filters to cope with oncoming headlights. Networks have been put in place with insufficient bandwidth to transfer the video generated by the installed cameras.
This results in significantly wasted investment. In the next chapter, we will discuss the right process to ensure that such waste is minimized.
Maximizing Return on Capital
The difference in price between the most effective product and one that barely works can be significant. But the difference in utility is infinite. If one feels that one cannot afford to purchase a product that can do the job effectively, it may make more sense not to purchase a product at all as ineffective products just provide an illusion of security and safety.
A well designed system can achieve multiple objectives using different types of sensors and can indeed achieve multiple objectives on each camera.The effectiveness of a product cannot always be measured in absolute terms.
If due to poor camera placement or a very difficult environment, a product can achieve 70% accuracy in a particular situation, it may still be far more accurate than the alternative which may be a human attempting to achieve the same objectives.
There are many components in a Smart City system and the intelligence software constitutes a very small proportion of the total cost. It is, however, the core of the system. It is critical to select the best available core as the return on the entire system can be compromised if the core does not work effectively.
The capital costs of the system are not the only costs that the users incur. There are many operational costs in using a sophisticated system as well.
If it takes a long time for the system to provide answers required by the user, it reduces his efficiency. If the system cannot function effectively when there is no one around in the control room, efficiency may be affected. If a large part of the system is not operational for any reason, this can affect efficiency. Reduced operational efficiency reduces the value of the system to the user.
iOmniscient has designed several features just to improve operational efficiency. The Jump to Event function is designed to answer the question about the time an event first started and about who was involved.
The Automated Surveillance Capability of iQ-Hawk provides immediate information about people and vehicles involved in an incident. Without such a system, operators may spend hours or even days attempting to extract the data from unrelated systems.
The Auto-archive function and the availability of information on mobile devices ensure that the system can continue to operate even when the control room is not manned.
The Automated Response Capability can enable the police and other emergency teams to arrive at the scene of an accident or crime much faster improving efficiency and possibly having an impact in a situation where few minutes decide between life and death for the people involved.
The iQ-Health Check system can advise the user if parts of the system cease functioning.
Every time the system is down and not operational, the user does not receive the value that he should expect from the system. Therefore, maximizing uptime is critical. To achieve this, the software and hardware has to be reliable and it is important to know when it is not working so that it can be immediately fixed.
Reliability comes from using good quality hardware and software. It can be enhanced through the use of redundant components.
Knowing that parts of the system are not functioning because they have broken down can be established fairly easily. The iQ-Health Check system can make the user aware of the system not functioning correctly even due to external factors, such as too much rain, a spider's web blinding the camera or movement of the camera due to vibrations.
The user can reasonably expect both the Systems Integrator and the suppliers to offer a level of service and ongoing support for fixing problems that is commensurate with the requirements.
The technology is advancing rapidly. Computing hardware and cameras can become obsolete and unusable in just a few years. Software improves even more rapidly and within a year, a system can be overtaken by new advances.
For this reason, iOmniscient offers an ongoing product update service to ensure that the system is continuously upgraded. In ten years time, the hardware may have collapsed and been replaced several times but the software should be as new as if it was bought that day.
The key to achieving this goal is that software must always be both forward and backward compatible with itself. It must be able to interface with former products while developing new interfaces for new devices and new products. This can only be achieved through strategic design. Very few products can achieve this goal. iOmniscient products are designed to be forward and backward compatible across versions because the objective is to eliminated obsolescence.
Throughout the book, we have emphasized the importance of openness and pointed out the negatives associated with products that attempt to lock users in with proprietary interfaces. This is much more important than attempting to reduce the risk of obsolescence. If a user is locked in to one supplier, it is impossible to take advantage of the latest improvements in technology. These are not the monopoly of one supplier. A commitment to open architectures and interfaces is the foundation to ensure that a system can take advantage of the technologies as they improve.