Video Analytics Cut Down Server Counts
Submitted by Aimetis | Date:
IP-based surveillance can lead to a greener environment and increased operational efficiency. Some benefits include reduced energy costs, increased server capacity utilization and higher scalability and availability.
Rising energy costs are quickly becoming the fastest growing operating cost, propelling corporations to embark on new conservation and "green" initiatives. Fortunately, green IT can reduce operational pressure, making it a win-win situation for both businesses and the environment.
The recent proliferation of video surveillance has increased its energy consumption dramatically, resulting in a higher TCO. However, the transition from analog deployments toward IP-based deployments helps surveillance become both greener and leaner.
Energy consumption from video surveillance is largely attributed to the electrical power required to operate servers. However, electrical power usage is not a typical design criterion for video surveillance, nor is it properly managed as an expense. This is likely because the billed electrical costs are not clearly linked to video surveillance, and the electrical bill is not within the responsibility or budget of the security group.
The fact that most video surveillance deployments today make inefficient use of the energy they consume is hardly a surprise. Industry research estimated that more than 50 percent of energy used in the server room goes to power and cooling equipment, rather than to computations and processing. Clearly, the true costs of running excess servers and storage are not well understood.
The Environmental Protection Agency recently reported that data centers in the U.S. consume 61 billion kilowatts per hour annually, or approximately 33 billion kilograms of carbon dioxide emissions, which is more than the total carbon dioxide emissions of several countries. According to an IBM Worldwide Study report, in terms of dollars, this translates into a total electric bill for data centers in 2005 of about US$2.7 billion and about $7.2 billion for the entire world. On a per server basis, this breaks down to an average of $550 per year for operating and cooling alone.
The best efforts should thus be taken to reduce the number of servers required for video surveillance deployments. To this end, IP-based surveillance offers clear advantages over analog deployments, with significant savings resulting from shared network infrastructure, computation and storage.
IP-based software such as video management systems (VMS) and video analytics plays an important role in reducing the hardware footprint, and important features to consider will be discussed below.
Camera Density per Server
Typical camera density ranges from four to 32 channels per DVR, and come in increments of four or eight channels. Consequently, many physical servers are necessary, since the total number of cameras exceeds one server. With IP-based software, more than 200 cameras can be managed and recorded using a single server, which normally takes about half the typical rack space of a 16-channel DVR.
The "one-to-one-to-one" hardware approach causes energy costs to increase. Each server hosts only one operating system that, in turn, runs a single application. Not only are these environments expensive, they are also the cause of IT inefficiency, with multiple servers operating at a fraction of their computing capacity.
This underutilization is typically seen with analog deployments. For example, supporting nine cameras may require the use of two 8-channel DVRs. However, both DVRs are operating with a CPU load of less than 10 percent. This "one-to-one-to-one" hardware approach also unnecessarily increases software maintenance and support, due to the sheer number of physical servers requiring software patches or hardware maintenance. Worse still, the risk of data loss is significant — not only is the cost of purchasing numerous backup servers prohibitive, the additional required rack space is nonexistent, and the additional 400 watts of electricity required to support them further adds to operating costs and carbon emission.
With virtualization, organizations can consolidate servers and easily reduce IT power requirements. For example, with only one physical server instead of two, server BIOS updates are cut in half, backups and disaster recovery becomes a fraction of the cost with fewer physical servers to protect, and network ports are relinquished with fewer servers requiring connectivity to the network.
Video analytics plays a sizeable role in reducing storage requirements by only recording when relevant activity occurs. This reduces space and energy requirements.
With analytics, video recordings caused by irrelevant background noise is greatly reduced, resulting in less disk space required for the same recording period. Better still, analytics reduces rack space and electricity dedicated for storage.
However, in many cases, separate vendors supply the analytics and the recording device. By adding an additional piece of hardware for analytics, which runs parallel to the existing recording product, the green IT strategy is undermined. This problem can be eliminated with a single software product for both VMS and analytics. Users can then take full advantage of the cost-effective benefits without increasing the number of servers for running analytics.
Green Licensing and Remote Surveillance
Most analytics and VMS vendors offer different license versions, depending on the surveillance requirements. The problem is that, more often than not, these versions cannot run on the same server. Licensing models should allow for available license versions to be mixed and matched on a single server.
Remote surveillance effectively eliminates frequent transportation to verify alarms and track site progress. Using analytics, monitoring personnel can be notified of suspicious activity, such as a perimeter breach, and verify the alarm before sending a guard to the site.
To put this in perspective, eliminating a 50-mile drive or a three-hour flight saves 88.2 and 1,543 pounds of carbon dioxide emission respectively.
Reducing Hardware Footprint
Using analytics to employ green business practices allows companies to reduce their customers' carbon footprint and contribute to the global environment in a positive way.