The Direction of VMS Technology

VMS technologies have come a long way over the past five years

• By Brian Carle
• Dec 01, 2017

Some of the industry trends which have affected technology and deployment design include a dramatic cost reduction in entry level NVRs and cameras. Merger and acquisition activity is reducing the number of independent technology vendors and driving popularity of single source technology providers and server hardware lines by VMS providers are now commonplace, allowing consumers to obtain the complete recording platform from a single vendor.

What might the next few years look like for VMS technology? Herein, we’ll explore trends and the direction technology may take for VMS platforms.

Information Security

Security is the current hot topic for video surveillance and other physical security systems, and for good reason. Several headline- making network attacks over the past few years have been made possible by security flaws in internet-connected cameras and DVRs. One notable, recent example is from late 2016. Dyn, a major provider of DNS services on the internet, suffered a sustained Distributed Denial of Service attack, mainly from compromised IP cameras and DVR/ NVR devices.

Emerging from these events, VMS vendors and security consumers see information security features as strategic and differentiating advantages. Some information security technologies which will become more commonplace VMS features.

• Encryption for system configuration files, communications, and authentication. System passwords and other sensitive configuration information is at risk when stored and transported. If authentication or configuration information is either intercepted during transport over the network or copied from a VMS client or server system, encryption will protect the contents, making them unusable to those without the encryption key.
• Encrypted video storage prevents video on a stolen hard drive from being readable. If a malicious actor has removed video storage drives after gaining physical access to a system, the recorded video will not be usable.
• Two factor authentication provides an additional verification step to validate that the user name and password entered matches the person attempting to authenticate. In the event a password is used by an unauthorized individual, the second authentication factor should prevent the individual from logging into the VMS.
• Encrypted export: Exported video is most at risk of being stolen. The exported video is often physically removed from the organization’s building, and there is no log of the access to the exported clips. If the media is lost, any individual can view the video, upload it to YouTube, provide it to the evening news, etc. As such, encrypting exported video ensures only authorized individuals with the encryption key will be able to view and use the export.
• Automatic system patching or in-application update reminders.

Co-processor Integration

Video surveillance is becoming a much more processor-intensive application. Many factors contribute to the increased demand for computing power, including increased deployments using H.265 compression, 4k camera resolution, and increased interest in deploying real-time and search based analytics.

VMS Client workstations require substantial computational power to decompress video, scale the video to the screen size, and display it. VMS recording servers may require substantial computational power to process video for motion detection, analytics, transcoding, or other operations.

To drive down the cost of server hardware and increase VMS performance, more VMS platforms will work toward integrating various co-processing technologies. Most common among these co-processing technologies are Graphics Processing Units, or GPUs, which are the processors located on a video card. When a VMS is integrated with such technology, some of the computing work will be processed by the CPU and some by the GPU. Adding in a graphics card with a compatible GPU will increase the computing power of the client workstation or server for much less expense than deploying a second computer or a more powerful computing platform that supports multiple CPUs.

Other co-processing technologies include Intel’s QuickSync technology as well as FPGA processors. Both of these technologies will likely be less common than GPU integration.

Leveraging Cloud

Interest in Cloud video surveillance deployments is growing. However, such deployments are frequently limited by bandwidth constraints. The upstream bandwidth available at a deployment site still largely dictates both how many cameras can be deployed, as well as the quality of the video available.

Despite scalability challenges, Cloud may still be leveraged by traditional VMS platforms in unique ways, creating hybrid Cloud or on premise deployments.

Traditional video surveillance deployments include hardware installed and owned by the consumer. The hardware will have a fixed amount of storage capacity, and consumers generally don’t specify more storage than is required due to the costs involved.

This traditional model is not very flexible. Often, a consumer’s video retention requirements may change or scene conditions may evolve, affecting how much storage is required to retain video. Consider the example of a school; storage requirements may be low during summer, but when school is in session, the additional scene activity drives camera bitrates higher, necessitating more storage space.

As such, VMS vendors may offer video archiving services. In-application options could allow consumers to pay for additional storage in the Cloud. When local storage is full, video would be uploaded to the Cloud for long term retention. Users can scale their video retention at any time without hardware changes under such a model.

In addition to leveraging Cloud for on-demand storage, cloud could be used for other “hybrid” applications, such as processing video analytics or simplified remote access.

In the case of processing for video analytics, recorded video could be uploaded to the cloud for processing and analysis, instead of having a server onsite to perform processing. This model could be useful in several ways. Take the example of a chain of retail stores: It may be desirable to process a camera’s recordings to generate a heat map, showing customer traffic patterns throughout the day. Uploading video to the Cloud for processing eliminates the capital investment in extra hardware needed to process the video. Additionally, if the retailer wants to process either a different set of, or more, cameras, on-premise hardware may not scale to the new demands. Using the Cloud, additional computing power can be obtained on-demand.

Finally, VMS providers could offer simplified remote access setup using a Cloud service. Remote access to a traditionally deployed video security system involves either using a VPN or setting up Port Forwarding and Dynamic DNS, both of which involve setup complexity and additional time for installation and configuration. Instead, a VMS provider could offer a Cloud service the VMS is preconfigured to connect to. The Cloud service would have a pre-established connection to the local VMS server which would allow the Cloud service to notify the VMS of requests for video from remote clients. This configuration would eliminate the need for port forwarding and DDNS or VPN setup because clients would not be making inbound connection requests to the on-premise VMS servers.

The End Result

Competitive differentiation and market demand constantly drive innovation in VMS products. More intelligent, secure, and flexible systems seem to be the current direction for enterprise VMS technology. These advancements will be enabled by co-processing technology integration, creative leveraging of the Cloud, and market demand for highly secure systems. The outcome will be systems that can do more and be easily expanded at lower cost for consumers.

This article originally appeared in the December 2017 issue of Security Today.