Using Video Analytics for Building Automation

Using video analytics beyond security may be the most important trend in video analytics.  While marketing and business intelligence applications have received the most attention to date, a growing interest exists in using video analytics for building automation.

ObjectVideo is the most vocal proponent of using video analytics beyond security.  Today, they released a brochure overviewing the business case for building automation. This report examines their positioning and explores the potential for this.

Benefit of Video Analytics for Building Automation

The core benefit claimed is that video analytics does a better job identifying when no one is occupying an area so that electricity and other utility use can be minimized. The ObjectVideo brochure focuses on the electricity reduction, estimating that the use of video analytics can save $1,100 USD per year in a 20,000 sqft facility by being more accurate than conventional motion sensors.

Can it Work?

One of the strongest benefits of video analytics in non-security areas is that the demand for accuracy is lower.  Using analytics for perimeter violation demands near perfect levels of accuracy.  However, for people counting or lighting regulation, 95% or 97% can provide material benefits.  Plus, these applications are indoors, far easier to achieve good results than the common fenceline deployment.

What's the business value?

The electricity benefits seem modest but material.  Let's take the $1,100 figure (OV uses $0.08kWh USD which is actually low given electricity prices in many places). Since this is for 20,000 sqft, you will need a number of cameras to achieve this.  It obviously depends on the layout of the building but I would estimate 10 to 20.  That results in savings per camera of $50 to $100 per year. However, if the layout has lots of open spaces (like a warehouse), the savings could be significantly more.

Because the savings are modest relative to the cost of the camera, you would need to use the building automation savings as part of a comprehensive package of benefits that video analytics can provide. The savings are not significant enough to deploy cameras by themselves.  This should not be a big problem because most facilities use a fair number of indoor cameras today.

ObjectVideo spoke with me and is promoting the use of low cost sensors that incorporate their video analytics.  To the extent that sensors (non video, like a motion sensor) are used, the cost could drop dramatically relative to using cameras.

How Can I Make this work?

• You need to run the analytics on the video.  This means either smart cameras or running this on a server (if you have existing analog or dumb IP cameras).
• The video analytics system will need to be capable of sending messages indicating when activity has both started and stopped.
• You need to send messages from the camera to the building automation system.  This will likely take time for different systems to add support for OV or other companies that might offer this.
• You do not necessarily need any intervention from your video management system.

What Does this Say About the Future?

While I think this while take a few years to reach common usage, this is a good indicator of the multi-functional world that video analytics will evolve to in the next few years.  To the extent that video analytics can provide modest but real benefits to a number of business activities, the stronger the case will be for the technology.  It will also simultaneously help to reducethe concerns many have in its use for security.