Contemplating the Cloud

The cost of storage, mobile access demands will define the future of hosting

• By Eric Fullerton
• Jan 03, 2011

Eventually, hosting will happen. And the video management software application will continue to have a central role in video surveillance because, by its nature, video gathers a lot of data input.

That data is going to be stored either on the camera, on a local server, pulled back to a central server or pulled into the cloud. But hosting is not going to change the importance of video management -- just the way it happens.

People are still going to want and need to continue to use all the great VMS features and functionality that let them work with their video data today.

It’s just a difference of how the data comes from the cameras into the user client of the VMS and how it is controlled and stored.

The Cost of Bandwidth
The cost of bandwidth is going to fall, and at a certain time when there’s enough low-cost bandwidth available, hosting will start to happen on a wider scale. The single biggest inhibitor for the hosting model today is the cost of bringing the video into the cloud. What happens when you bring the video surveillance into the cloud is that it adds more cost to the stored video in a surveillance system. This indirectly and negatively impacts the total cost of ownership and return on investment calculation for operating a video surveillance system.

Some people say that not more than half a percent of all video is ever viewed after it is recorded; others think it could be as much as 5 percent. So if hosting is to bring all this video into the cloud for storage, in reality you are adding cost to 100 percent of the video when you are interested in only a maximum 5 percent of it. So when you are using up to only 5 percent of that video, the accumulated cost of the total storage in the cloud would be prohibitive at this time because only 5 percent of it has to carry the cost of 100 percent of it.

One scenario to improve this cost dilemma could be to have only smart cameras -- with analytics on board to define which video segments need to be sent to storage. But if you look at the nature of the way people use video today, they always like to have access to whatever was recorded. Therefore, there might be some improvements doing local recording at the edge to lift only a certain amount of the video into the cloud for later viewing. Still, that’s not the way people are using video now.

The Tipping Point
The real enabler of hosting will be the day when the cost of using a hosted solution is equivalent to the cost of doing local storage and centralized management in private networks -- this is very similar to what enabled hosting for the IT data world in the 1990s.

There was a cut-over phase when the cost of putting the data into the cloud became lower -- that started fueling the whole cloud computing trend, where the tipping point will be for video to be cost efficient in the cloud. We don’t know yet.

Mobility is another driver that will influence hosting. Devices such as smart phones and tablets are significantly changing the way people do things. But that’s just on the user-client side, in the way people want to access their data.

For video data, we’ve already enabled remote access through the Internet or private access through VPNs without going into what is defined as hosting. So mobility is driving a new way of accessing and using video, but it is not necessarily driving true hosting in the cloud. It could be one of the elements that have to be in place for hosting to happen -- as people expect to have full access to full-featured video at any place and any time.

As we move forward, there will be video stored in cameras, in local servers, in central servers and, eventually, in the cloud. The demand on a video management system in the future will be to give people seamless access to whatever video there is at the point of a fingertip with maximum usability.

This article originally appeared in the January 2011 issue of Security Today.