WITH benefits such as superior image quality, enhanced video analytics and the integration of surveillance video with business intelligence software, the tipping point at which IP video overtakes analogue is rapidly approaching, and the comparison of their total cost of ownership TCO is becoming irrelevant.
Regardless of which system type ultimately costs more to implement, IP video has some clear cost advantages in its own right. Examples include the ability to run video over a converged cabling infrastructure, using Power over Ethernet (PoE) switches to provide centralized power to cameras, and flexible implementation options (such as wireless video). 
That being said, the switch to transmitting video over the data network can be significantly more complex than with analogue systems. 
Understanding the true TCO of IP-based video surveillance systems can be significantly affected, particularly when assessing the impact on your organization’s Information Technology (IT) systems. 
A number of unpleasant surprises may lurk for those who fail to take a holistic approach when assessing costs of an IP-based system. All too frequently, IP systems are still designed by security specialists who may know analogue surveillance systems cold, but are still developing their understanding of both IP data flow and IT operations.

“If your organization has committed to making the leap to IP video surveillance technology, understanding the true TCO of your new system will require careful attention to details regarding your data network, facilities and IT operations.”

In many cases, initial assessment of the impact of surveillance on the existing technology environment can be minimalistic, if not outright dismissive. Do the following network analyses sound familiar to anyone?
“You have an IP network already and you can use your existing data cabling to save money on implementation. Just put in some network switches that support PoE, make sure multicast is enabled and you are good to go.” 
“Bandwidth? Not a problem. With H.264, your megapixel cameras will only require 6 Megabits per second each, and all of your switches support Gigabit.”
Over-simplified arguments like this can lead to significant cost oversights, as even a moderately-sized IP-based video system can greatly impact network performance and add substantially to IT capital and operational costs. If your organization has committed to making the leap to IP video surveillance technology, understanding the true TCO of your new system will require careful attention to details regarding your data network, facilities and IT operations.

The basics of the IP video TCO

At face value, the hard costs of implementing an IP-based video surveillance system are straightforward. Major cost components of a new system typically include:
* System design
* Cable installation. While IP video runs on a converged cabling infrastructure, new cabling usually has to be installed from the nearest data closet to camera locations
* The IP cameras, including mounting and maintenance costs over their projected lifetime. For many systems, this also includes software licensing for each camera
* Power and data distribution, which typically includes PoE data switches and possibly Uninterruptible Power Supplies (UPS). This also includes warranty and maintenance costs
* Servers, network storage and software. This includes maintenance service contracts, and frequently includes annual licensing costs for the software and for each camera
* Viewing stations, including video walls. This also may include annual licensing costs for software
* Training costs, both for users of the system and the IT staff who will maintain the software
* Capital costs for data center and data closet rack space, as well as the overall equipment footprint.
Although these components are generally understood to comprise the majority of costs, there are some common oversights. Many organizations struggle with accurately assessing the impact any new technology system has on energy costs, and surveillance is no exception. Beyond a general assumption that PoE can lower capital costs for providing power to cameras, power consumption is frequently not considered in the cost of a new surveillance system.
Given the centralized nature of video processing and storage for IP-based surveillance solutions, energy costs of a new system can be significant, particularly in the data centre. In addition to the ongoing power consumption of servers and storage, the power consumption for air conditioning systems in the data centre can also be adversely affected. 
Upon completion of system design, a table-top power review that includes both power draw and heat dissipation requirements can usually be performed quickly and easily, giving organizations a more complete view of the new energy costs that will be acquired with the new system.
Then there are software support costs. Video servers and viewing stations may require antivirus software that requires annual license payments. Furthermore, many IT departments have a standard suite of software for monitoring and maintaining servers and network components that will require additional licenses. In addition to these costs, new servers and client applications may also require significant and ongoing IT labor to assure underlying operating systems are monitored and patched on a regular basis to assure system integrity. 
Don’t forget video backup costs. For organizations that are required to retain video for any length of time, new systems will typically include a backup system for archived video. If your IT department has an existing backup system, it may be leveraged for video storage. However, for even a moderately-sized surveillance system, the existing backup system will typically need to be extended. For tape-based systems, the cost of additional tapes and magazines may also be significant.

Hidden pitfalls of TCO projections

In addition to common cost oversights described above, there are many hidden factors that can adversely affect the TCO of IP-based video systems. Some of these factors that can wreak particular havoc on cost projections include things like camera creep. 
The rapid advancement of IP camera technology can cause headaches for organizations and integrators alike when these advances exceed the pace of system implementation. Integrators may specify a given camera during the bidding process, but by the time the infrastructure is built, it’s likely that newer, higher performance cameras are available. If the customer changes out the camera selection, the integration team must analyze the effect that change has on the entire surveillance system and the network – everything from switching to storage to the horsepower of viewing stations.
To the extent possible, the system design should include realistic projections of the maximum capacity the video surveillance system may require over time, both in terms of the maximum number of cameras, and their potential network and storage requirements. The implementation plan should also incorporate specific timeframes for purchasing system components, with appropriate check points to assess the impact of specification changes for one system component on other subsystems.

Network quality of service

In an optimal configuration, the simple network math discussed earlier works. If the installed switches support Gigabit Ethernet, how can a few cameras that consume 6 Megabits per second each possibly choke the network? The truth is that while surveillance video may flow across the network just fine in this situation, unfortunately many surveillance implementation projects are designed in blissful ignorance of other network traffic that can be severely impacted by the new system, or how existing traffic can affect the surveillance video. 
Many organizations have network traffic that is much more sensitive to latency (network delay) than surveillance video; examples include Voice over IP and video conferencing traffic. Many existing networks are not configured appropriately to assure the quality of service for this traffic, which can go unnoticed until the addition of IP surveillance to the network degrades the quality of voice traffic, which can directly affect business operations. 

“If your IT department has an existing backup system, it may be leveraged for video storage. However, for even a moderately-sized surveillance system, the existing backup system will typically need to be extended.”

Degraded voice traffic may also cut both ways, as audio is increasingly being added to IP cameras and can be much more sensitive to latency than the surveillance video itself. If network Quality of Service (QoS) is not configured prior to the introduction of video, IT managers may be put in a position of requiring an expensive reconfiguration of the existing network on very short notice, while suffering service interruptions in critical voice and data systems until the upgrade is complete. 
Prior to implementing an IP-surveillance system, review the current network configuration to assure Quality of Service (QoS) is implemented appropriately to assure integrity of existing services, as well as assure the network can meet the new demands of surveillance video. Also, identify potential bottlenecks where video traffic may be consolidated, such as clusters of viewing stations located on the same edge switch(es). 
These points on the network may require more robust data throughput than is currently available. This is particularly important in the data centre, where network engineers should carefully review not only camera to server traffic, but inter-server traffic, including traffic flow between storage and backup systems. For large surveillance systems, the need to upgrade core switch capacity can be one of the most expensive surprises of all.

Optimistic Storage Projections

While network storage costs continue to decrease over time, storage can still be the largest capital expenditure for an IP-based surveillance system, particularly if an organization has a large number of high-quality IP cameras recording at full frame rate, and/or has long retention requirements. Many manufacturers provide an array of technologies and techniques to optimize storage, but be vigilant when storage projections are based upon predictions about the nature of the video your organization will capture, particularly if the surveillance system is new.
For example, some manufacturers support the capability to automatically control the camera frame rate based upon video activity, transmitting and storing one frame per second when there is no motion, and increasing the frame rate to full-motion video when activity is detected. Technologies such as this can drastically reduce network storage and bandwidth requirements. 
However, projecting the actual storage reduction requires making an “educated guess” as to the amount of motion that may be expected from each camera based upon their location and view. It can be tempting to overestimate the benefit of such technologies and their impact on storage costs, resulting in delayed sticker shock when an organization realizes months after implementation that their storage predictions were wrong. The sticker shock frequently extends to both tape-based and disk-to-disk backup systems that may also require expansion.
For this reason, organizations should carefully review all of the underlying technology assumptions that factor into storage cost projections, and have a complete understanding of the potential costs if the assumptions are incorrect. 

Power, wireless, security

Among a number of vital considerations are incomplete power assumptions. To the extent that organizations consider power consumption of their new system, most efforts to assess energy costs are confined to the data centre. The costs of assuring appropriate power and backup power in data closets can also be significant, particularly for organizations that use PoE to power additional network devices such as wireless access points and IP phones. 
For safety and compliance reasons, many organizations have specific standards for UPS backup power to assure PoE switches that support IP phones will supply power for a set duration in the event of a power outage (typically 30 minutes). When IP cameras are plugged into PoE switches in the same closet, the impact on the UPS infrastructure can be significant, as IP cameras can consume substantially more power than typical IP phones.
Next comes the possibility of wireless video options. The majority of current surveillance systems live on wired networks. However, with rapid advancements in surveillance technologies, organizations are increasingly eying extending surveillance systems to places where placing wired cameras can be cost-prohibitive, such as parking lots, industrial facilities, and entire cities.
While wireless surveillance technology can provide significant cost savings, it is critical to pay careful attention to the design of the wireless network that will support surveillance video, particularly when live video is continuously monitored. Bandwidth of wireless technologies, particularly those that use unlicensed frequencies, is always theoretical. 
A wide array of factors can affect a wireless network’s capability to transmit live video, many of which are beyond an organization’s control. 
With careful consideration it is possible to build robust, resilient wireless networks to support video; this resilience typically comes at a price premium. Avoid cost-driven temptation to under-design the wireless network, and have a pre-defined plan to adjust wireless signal quality, throughput capacity and QoS as required.
Then there’s security complexity. For many organizations, the flexibility of IP-based systems can be accompanied by significant security challenges that affect operational costs. Analogue video systems that are closed in nature are relatively simple to secure, and typically include simple tools to assign access permissions and log system activity. 
For IP-based systems, these security controls move into the application layer where it can be difficult (and expensive) to configure security appropriately, particularly for organizations with strict compliance requirements. When designing a new system, carefully review the data security design against not only your organization’s security and compliance policies, but against your IT policies as well. Take special note of operational costs to maintain the security and integrity of the data continuously throughout the system’s information lifecycle.

Strategic planning to minimize TCO

Maximizing your return on your investment in an IP-based system requires careful planning to minimize implementation and operation costs. When developing a strategic plan be sure to plan together with other stakeholders. Developing a complete understanding of TCO will require expertise in security, facilities and IT. 
These functions are under separate management in many organizations, and contention and confusion over supporting roles for technology are unfortunately common. As the underlying technology supporting all of these functions marches towards a converged infrastructure, it is critical to leverage the expertise of all of your internal resources to assure successful implementation of your new system.
Create a vision is also important. Unexpected costs frequently occur because the business benefit of an IP-based video system generates organizational demand to expand the system far beyond its original design. When planning a new system, carefully consider the potential for full expansion of your system, even if the initial implementation is modest. It is important to assure that your system is scalable to meet your future needs, whether or not the system is ultimately expanded.
Finally, look ahead. IP-based surveillance technology is still evolving rapidly. Periodically regroup to seek and evaluate new IP surveillance technologies throughout your system’s lifecycle. When new solutions become available, determine if it they provide sufficient business value to incorporate them into your surveillance system. If so, keep a wary eye on their cost impact to your technology infrastructure to help your organization assure a positive return on your business investment in the power and security of your new IP surveillance system.
*Patrick Luce, director of consultative services, Vector Resources