JASON Calacanis, the founder of the people-powered search
engine Mahalo recently claimed a new online storage service is coming from
Google. What’s more, according to Calacanis, the service will be free. It’s a
development that could have ramifications for CCTV people further down the
track.

“Tried to get google rep to give launch date for
Google Backup service,” Mr Calacanis said in his online update late last
month. “He got mad flustered … u can be sure Google Backup is
coming.”

And Calacanis underscored his speculation at the recent
CEBIT show in Sydney.

“From what I know it’s coming out,” Calacanis told
the press. “I know people, and I’ve heard that it’s being worked on.”

Simply stated, an online backup services allows users to
save their files to offsite locations in data centres featuring robust
intrinsic design features like HDD mirroring, dual servers, duel power supplies
and super clean and stable environments. It’s been estimated that within 10
years up to 80 per cent of PCs will back up to off-site locations.

The importance of this Google development from the point
of view of security electronics and networking people is its potential to popularize
MPLS cloud network architecture. This is a model that has the potential to
change networked video surveillance solutions significantly. True edge devices including
CCTV cameras from manufacturers like Mobotix could store video in onboard NAND
bins and upload events to data centres overnight.

The big deal here is cost. At a certain price threshold,
use of data centres as a form of primary storage becomes compelling –
especially for small/medium businesses or larger organizations which have
significant existing investment in data centre rack space for general network
support.

ISPs tout data centre storage costs as having dropped to
the point where off-site storage is now a commodity but in truth this model
remains some way off and that’s why Google Backup, if true, is so important. It
will take a broad shift in the market to induce the lowered upload costs that would
make offsite storage attractive for high demand metadata like video. More
customers, means more competition, means lower prices.

There are some technical challenges ahead, too. As
megapixel cameras become more prevalent, network loads and storage demands are
only going to increase. Manufacturers and end users will address these demands
in a number of different ways. Cameras will incorporate H.264 compression which
reduces file sizes by around 70 per cent for a given resolution. They’ll also
look to incorporate analytics that allow cameras to record movements rather
than blithely recording a scene for days and days without change.

And end users may need to adjust their expectations. For
many applications there is no need to record scenes in which there’s no
movement. Another potential change is going to relate to recorded frame rate.
We all love the idea of real time recording but most megapixel manufacturers
are talking much less. Some seem to have settled on 15 images per second while
others are preaching as few as 4.

Ultimately, frame rate selection depends on application
but recording all cameras regardless of non-event, at 30 frames a second and on
all inputs is an inefficient use of resources.

Most network people envisage a long term future in which
all systems reside on clever, robust wide area networks supported by off-site
data centres. What’s going to be interesting is just how the electronic
security industry finds its way to onto those networks.