Learned Memories

In the years since the 9/11 attacks, concerns about terrorism and crime have driven the use of security systems and surveillance cameras into areas and markets that might not otherwise have considered such technology necessary. This spread into new markets has led to evolving technologies that fill the demand for more intelligent and automated security systems. Behavioral recognition is one such emerging technology that should already be on electrical contractors’ radar.

Behavioral recognition systems (BRSs), or video analytics-enabled systems, constantly examine camera images to study people and vehicles and determine whether they are a security threat.

“This is a rapidly emerging market with significant growth potential,” said Chris Brown, executive vice president of Check Video, Reston, Va.

According to Forbes, the reason for the quick adoption of cameras that use behavioral recognition technology is that people are not good at attentively watching large amounts of video for very long and, with an estimated 30 million surveillance cameras in the United States, that means more than 4 billion hours of footage every week.

“Video analytics-enabled systems use the images captured by video cameras and marry them to intelligent algorithms to determine movement and the presence of vehicles to assess threats in real time,” Brown said.

Prior attempts to create rule-based video analytics to verify behaviors have not been very successful, said Hobby Wright, vice president at BRS Labs, a software company headquartered in Houston.

“There’s been a decade’s worth of experiments in video analytics that have taken video; applied rules to it, such as geographical area, object types, direction of travel, time spent in defined area, etc.; and tried to determine security threats. But BRSs do not require rule development, but rather machine-learning algorithms that enable the system to couple with video analytics and provide an intelligent solution that actually learns, much like the human brain, to recognize suspicious behaviors and alert the appropriate people in real-time,” Wright said.

In effect, the system forms memories, builds models of them, and alerts security staff of unusual behaviors that do not conform to “learned memories.”

The end-user determines the desired activity parameters inside the camera range, as well as the areas and schedule within which to have detection. The system then sends a 10-second clip to appropriate personnel upon detecting activity that meets those parameters.

“The end-user can also set the system up to coordinate with the alarm panel so that, when an alarm goes off, the camera generates a video clip of the event,” Brown said.


Municipalities and law enforcement are no longer the only ones who have taken note of this technology.

“Retail, commercial office and multi­family and low-incoming housing markets are seeing an enormous rush of new customers to this technology,” Brown said. 

These markets, with mostly high-­density populations, have traditionally used guards for security. Intelligent video can help these building owners reduce the number of guards needed to protect the space or, in partnership with remote monitoring services, can eliminate the guard staff along with human error as well as reduce security costs. Construction sites are another strong market for this technology, Brown said, because intelligent cameras with video analytics-enabled software—which can scale to an unlimited number of cameras—eliminate the need for fences, and the cameras are portable and can be moved as construction progresses and develops.

ECs can promote this technology as an effective way to use cameras to protect assets and people. In addition, intelligent video surveillance systems open a large opportunity for recurring revenue for the contractor that did not exist before.

“Contractors have the opportunity to associate the video to a monitoring service for the customer’s security, business optimization, and remote guard services and negotiate a percentage of the monthly fees paid to the central monitoring company,” Brown said.

Contractors must understand that their customers might be skittish at first about the amount of artificial intelligence inherent in BRSs.

“Those contractors that become familiar with the technology, however, can demonstrate its effectiveness and allay any fears,” Wright said. 

Customers might also be wary of the price of such intelligent cameras, but what cost up to $1,200 three years ago now can be found in the $500 range, depending on the quality of the camera and its detection software.

“Contractors that thought the market too costly few years ago, should certainly re-examine it now,” Brown said.

In addition, industry organizations such as the Central Station Alarm Association are beginning to write standards for video as a service and are addressing privacy issues, such as the use of video in court and in the work environment.

“However, as long as the system user owns the property and has adequate signage warning that video recording is in progress, liability concerning privacy is mitigated,” Brown said.

About the Author

Darlene Bremer

Freelance Writer
Darlene Bremer, a freelance writer based in Solomons, Md., contributed frequently to ELECTRICAL CONTRACTOR until the end of 2015.

Stay Informed Join our Newsletter

Having trouble finding time to sit down with the latest issue of
ELECTRICAL CONTRACTOR? Don't worry, we'll come to you.