Living on the Edge

Video recording on the edge has been available to security practitioners for some time, but never with as many affordable and easy-to-deploy options as today. And since the video surveillance market continues to grow more rapidly than any other segment of physical security, edge recording is quickly becoming a mainstream application.

In the past, most CCTV video surveillance systems were fed video over coax cable from analog cameras, displaying it on control room monitors, while VCRs recorded to videotape. More recently, VCRs have been replaced in new or upgraded surveillance systems with one or more of three choices: DVRs with analog surveillance cameras, NVRs with network surveillance cameras, or hybrid DVRs with both analog and network cameras. In many cases, although the recorder itself has changed, much of the rest of the infrastructure design has not, and video continues to be centrally recorded.

Edge recording techniques are the next major step toward more cost-effective use of video surveillance systems, enabling those systems to be more valuable to the organization.

Recording on the edge requires several technologies to be truly effective. Edge recording is not just using an Ethernet-based IP network and adding smart network surveillance cameras; nor is it only plugging in IP video storage for higher reliability and scalability. Instead, edge recording uses all of those technologies and more, functioning together in a comprehensive solution.

What is Edge Recording?
The premise of edge recording is simple. Anything that is recorded and stored at an individual “edge” location, such as an office environment, factory floor, restaurant, highway or city street, parking lot or retail location, doesn’t have to be immediately uploaded to a central storage and monitoring location. It can instead be recorded and held locally, referred to as the edge.

In fact, using a technique called “store-and-forward,” video can be held at the edge and only transmitted over the network as required via a set of preset criteria or conditions. This is important because one of the major challenges facing users of IP network cameras is that while they bring many advantages over their older analog equivalents ranging from night and low-light vision, digital zoom and Power over Ethernet (PoE) support, they use the IP network infrastructure to transmit their video capture to a storage device.

Network Bandwidth
Multiple IP network cameras can plug directly into a new or existing network quite easily; and IP network cameras can produce a non-stop, constant stream of megapixel-quality video. This streaming, however, can consume a lot of the network’s bandwidth, or its overall capacity.
When multiple surveillance cameras stream video over a corporate IP network, there is often little bandwidth available for other cameras or for any IT applications normally using the network.

Corporate IT is often unwilling to allow video applications to run over their existing Ethernet network, due to two concerns: bandwidth consumption and network security. Network security issues are real, but can typically be worked out between physical security and IT through qualification and certification. Network bandwidth consumption can be a much larger issue and trickier to solve.

However, by keeping IP video streaming largely to an edge environment, instead of consuming all available corporate network bandwidth, IT concerns can be alleviated.

Smart Cameras and Video Analytics
To further lessen storage and bandwidth needs, event-based recording is growing more common for video surveillance users.
By employing video analytics and sensor suites embedded into network cameras, or in add-on intelligent video servers, event-based recording can greatly reduce the amount of storage to be retained, stored on the edge and eventually transmitted to another location. An entire ecosystem of surveillance camera makers and analytics vendors fulfill these “intelligent video” needs, and offer a wide range of products.

As an example, by using sensors to detect movement, heat, sound or other predefined conditions to start and end recording within a preset time period, the amount of storage required can be greatly reduced, or additional IP surveillance cameras supported with the same storage capacity.
Advanced algorithms in the camera or video server can be used in real-time to monitor for pre-identified conditions and classify them accordingly. If a threat or other event to be recorded is identified, operators can be alerted while the event is occurring and the video transmitted to one or more locations while recording continues. This can be accomplished largely by automation and without constant human monitoring, and thus this intelligent application can be the key to a successful edge recording solution.

This is particularly important when organizations use large numbers of network cameras in one or more locations. Such an environment can threaten to overwhelm operators with information overload, and lead them to inadvertently overlook genuinely important events. By automating basic analysis activity and combining it with the forwarding of only those events with crucial information, the amount of video traffic on the IP network can be dramatically reduced.
This has the added benefit of improving the value of the video transmitted for central recording and often results in the improved performance of the operator team. Finally, some video analytics solutions compress and limit the number and quality of frames transmitted for retention, further reducing both edge and central storage needs.

IP Storage on the Edge
Beyond reducing the amount of video that needs to be recorded and forwarded to central or multiple locations for action or retention, another very useful edge technology is IP video storage.

Video surveillance has unique workload characteristics that can often result in PC-grade disk drives failing as they age. Unfortunately, these same PC-grade drives are typically found in many DVRs, hybrid DVRs and NVRs today. The resulting failures often lead to lost video recordings, with the DVR or NVR being down and recording stopped until the failed drive or entire system is replaced.

While enterprise-grade IT storage solutions typically use higher-quality disk drives and include advanced data protection features not common in DVRs and NVRs, they are not well designed for video workloads. IT systems are designed primarily for database, e-mail and general office applications, and not the streaming workloads required for video storage.

Unfortunately, enterprise-grade IT storage is also much too expensive for the physical security industry, and security practitioners in the past were often left only with lower-quality solutions that could not support video workloads or scale sufficiently to be truly effective.

The solution is to use specially designed IP video storage, architected for use on the edge. Such edge video storage can be connected to DVRs, NVRs or video servers for an immediate boost of storage capacity and reliability. IP video storage can also be connected directly into the local IP infrastructure with network cameras, and with or without DVRs and NVRs.

Edge solutions using IP network storage do not need to be expensive. The Manufacturers Suggested Retail Price for high-performance, highly reliable IP network storage should be well below $2,000 per TB (a TB, or terabyte, is 1,000GB) for standard configurations, and slightly above that for fully fault-tolerant deployment needs.

Many existing DVRs and NVRs can be easily and quickly upgraded by security integrators to connect this specially-designed edge video storage, increasing the retention capacity of the recorder in just a few minutes. Analog and network surveillance cameras, cabling infrastructure, operational procedures and even video management applications remain unchanged when the proper edge IP video storage is used.

Adding days, weeks or months of recording on the edge allows these IP storage solutions to greatly reduce the need to upload video. IP video storage retention capacity is much greater than most DVRs and NVRs, even at the highest frame rates and resolution.

The best edge IP video storage also delivers advanced data protection. Since most DVRs and NVRs use not only PC-grade disks but also include little data protection in their design, video recording can be permanently lost when disk drive failures do occur.

This risk of lost video recordings can be eliminated by adding IP video storage to the edge, with advanced RAID (Redundant Array of Independent Disks) technology and fault-tolerant components. These features help ensure unmatched, non-stop recording so that unlike with unprotected DVRs and NVRs, even if a disk drive does eventually fail in the IP video storage, recording continues and video is no longer lost.

Deploying Edge Recording Solutions
While adding just about any of the technologies that make up an edge recording solution will have benefit, only when deployed in an integrated edge solution will the maximum gains be realized.

The place to start is with your security integrator, selecting one who is not grounded exclusively in the CCTV designs of the past, but who understands the benefits IP and edge technologies bring for their customers. You will find this is an increasingly large list, as more and more security integrators learn about and deploy this technology, on the edge and elsewhere for improved video surveillance.

Jeff N. Whitney is vice president of marketing at Intransa (, a scalable IP video storage vendor. He offers extensive physical and network security knowledge, combined with IT storage expertise developed over 20 years as an industry practitioner. Mr. Whitney is a volunteer to the ASIS Intl. Physical Security Council, and is the primary Intransa representative to the Security Industry Association, while also representing the company to The and the Storage Networking Industry’s Green Initiative. He can be reached at